Platelet 5-HT_2A receptors

A short premise should be made herein. In pharmacology, the binding parameters are expressed as maximum binding capacity (B _max, corresponding to the number or density of binding sites), and constant (K _d), that is reciprocal of the affinity constant (K _a), and indicates concentration of the drug necessary to saturate 50% of all the sites present. Furthermore, the Michaelis–Menten equation, generally, employed to assess the enzyme kinetics, is also used to calculate the Michaelis–Menten constant (K _M) of the reuptake, while representing the substrate concentration necessary for the reaction to reach a velocity equal to half the maximum one (or semi-maximum speed). The rate of reaction when the reuptake proteins are saturated with substrate represents the maximum rate of reaction (V _max). The relationship between rate of reaction and concentration of substrate depends on the affinity of the ligand for its substrate.

Out of the total 16 studies selected, nine were longitudinal and seven cross-sectional.

Longitudinal studies, assessing receptor affinity parameters at baseline and after drug treatment, reported controversial results. At baseline, five studies showed no change in B _max, ^138–142 three increased B _max, ^143–145 and one decreased B _max,Reference Markovitz ¹¹⁷ with no changes in K _d value.

At the final endpoint, no change was observed in B _max in five studies,Reference Sheline ^{139 ,} Reference Bakish ^{140 ,} Reference Stain-Malmgren ^{142 ,} Reference Pandey ^{144 ,} Reference Hrdina ¹⁴⁵ increased B _max in three studiesReference Markovitz ^{117 ,} Reference Cowen ^{138 ,} Reference Alvarez ¹⁴¹ and decreased B _max in one study.Reference Biegon ¹⁴³ Furthermore, one study with 6 months of sertraline treatment showed a decreased K _d.Reference Markovitz ¹¹⁷

In one of the first studies, no statistically significant differences emerged in B _max or K _d in a sample of 26 patients with a diagnosis of major depressive disorder (MDD) and 25 age-matched controls at baseline. In this case, the 5-HT_2A binding properties were evaluated using the [¹²⁵I]-LSD. However, after 6 weeks of tricyclic antidepressants (TCAs, amitriptyline, dothiepin, and desipramine) B _max values resulted significantly increased.Reference Cowen ¹³⁸ It should be highlighted that the impact of TCA treatment on the binding parameters was tested only on 11 subjects. Furthermore, it is very difficult to ascertain the possible pharmacological characteristics of each compound (amitriptyline, dothiepin, and desipramine). In another longitudinal study, comparing depressed placebo responders (PR = 10), depressed placebo nonresponders (PNR = 23), and controls (C = 14), no statistically significant differences at baseline and at final endpoint were detected using [¹²⁵I]-LSD as ligand.Reference Sheline ¹³⁹ The results of this study could be influenced by the numerical intergroup heterogeneity and by the short observational time (10 days). In an 8-week, double-blind trial with two selective 5-HT reuptake inhibitors (SSRIs, paroxetine versus fluoxetine), using [³H]-LSD, no changes in B _max were found at baseline and at final endpoint. Nevertheless, dividing the depressed group into suicidal and nonsuicidal, the B _max resulted higher at baseline in suicidal depressed.Reference Bakish ¹⁴⁰ These findings should, however, be interpreted in the context of some limitations: the relatively small sample size and the treatment with two different drugs. In a sample of 54 subjects (27 unipolar depressed [UD] and 27 controls), using [³H]-LSD, in a 12-week clinical trial, there were no significant differences at baseline. However, a statistically significant increase in B _max was observed after clomipramine treatment.Reference Alvarez ¹⁴¹ The small number of patients represents the main limit of the study. Similarly, Stain-Malmgren et al,Reference Stain-Malmgren ¹⁴² in a longitudinal study with a sample of 30 MDD patients and 30 controls, described no baseline intergroup differences in B _max and K _d ([³H]-LSD) and no changes after 6 months of treatment (paroxetine or sertraline). It should be stressed that this is a long observational study compared with the previous ones, and that the many outpatient check-ups led to a high therapeutic compliance.

In a small sample of 12 unmedicated depressed patients, the mean B _max value was twice as higher as that of 12 age- and gender-matched control individuals. After 4 weeks of antidepressant treatment, B _max values were normal. The 5-HT_2A-receptor binding was determined using [³H]-ketanserin. The authors hypothesized that the increased B _max might be an adaptive mechanism due to a reduction in 5-HT availability, readily restored by drug treatment.Reference Biegon ¹⁴³ However, given the small number of patients, it was not possible to distinguish the effects of clomipramine and trazodone.

Others reported similar findings at baseline with [¹²⁵I]-LSD in 23 depressed patients, underlying how the increase in B _max was more represented in those depressed subjects with a recent history of suicide attempts or suicidal ideation. No correlations were identified between B _max, K _d, illness severity, and drug response. In this study, no changes were detected after 4 to 6 weeks of treatment.Reference Pandey ¹⁴⁴ However, it must be said that the authors did not specify the drugs used and did not show the B _max and K _d values. Another 8-week longitudinal study, comparing 60 UD with 40 control subjects, supported the increased density of 5-HT_2A receptors, calculated with [³H]-LSD, mainly enhanced in women and in the suicidal subgroup. According to these authors, the persistence of this difference after the antidepressant treatment would suggest that 5-HT_2A up-regulation is a trait rather than a state-dependent marker.Reference Hrdina ¹⁴⁵ Nevertheless, it should be noted that the authors treated the depressed subjects with different antidepressant classes (SSRIs, TCIs, SNRIs, placebo, etc.) without analyzing them separately. Only one study, comparing 21 MDD patients and 21 age-matched control subjects, showed a reduced B _max at baseline using [³H]-LSD as a ligand. In this case, sertraline treatment not only restored receptor density of 5-HT_2A but also reduced K _d.Reference Markovitz ¹¹⁷ The small sample size, the short duration of treatment (6 weeks), and the lack of a placebo control group represent the main limitations of the study.

Even the cross-sectional studies report conflicting results: four studies do not describe statistically significant differences between the depressed and the control groups regarding B _max ^146–148 and three studies showed an increase in B _max in depressed subjects. ^149–152 The receptor affinity showed no differences between the two groups, except two studies reporting reduced K _d value.Reference Sheline ^{151 ,} Reference Rosel ¹⁵³

By employing the [³H]-LSD ligand, a sample of 30 MDD patients and 30 age and matched controls showed similar B _max and K _d, as well as a positive correlation between anxiety rate, MADRS scores, suicidal behaviors, and B _max.Reference Neuger ¹⁴⁷ The same results were obtained again with [³H]-LSD in 88 MDD subjects (of which 60 with suicidal behavior and 28 nonsuicidal) and 123 controls, with no difference between patients with and without suicidal behavior had no differences compared with non-suicidal depressed patients.Reference Roggenbach ¹⁴⁸ No difference in B _max was also found in other two studies. In the first, 5-HT_2A receptors were labeled with [¹²⁵I]-LSD and one group of 43 MDD patients was compared to 42 healthy controls.Reference McBride ¹⁴⁶ In the second study, [³H]-ketanserin was used and 51 MDD patients were compared with 31 control subjects.Reference Rosel ¹⁵³ It should be noted that, in all these studies, patients and healthy subjects were not well matched, while being quite heterogeneous in terms of sex, age, and size.

One study labeling 5-HT₂ receptors using [³H]-LSD, reported increased B _max values in a sample of 29 depressed patients, which are more evident in women.Reference Arora and Meltzer ¹⁵⁰ The depressed group was not homogeneous from a psychopathological point of view, as it included subjects with BD, UD, and schizoaffective disorder experiencing a current major depressive episode (MDE). Furthermore, it should be added that the groups were not well age- and sex-matched. Confirmations of the increased density of 5-HT_2A came from two different cross-sectional studies comparing UD and control subjects. In one study, 25 depressed patients, especially women, showed also increased K _d values with the receptors labeled by([³H]-LSD.Reference Hrdina ¹⁵² In the second case, using the [¹²⁵I]-LSD, a group of 35 MDD patients showed a positive correlation between the cognitive symptoms of depression and B _max.Reference Sheline ¹⁵¹

As already underlined, most of the findings are probably affected by methodological limitations, such as the neglect of psychiatric or medical comorbidity, or of confounding factors in the evaluation of single serotonergic parameters (such as cigarette smoking), differences in the diagnostic evaluation (considering, eg, different affective disorders belonging to distinct psychopathological spectra as a unique depressed group), the severity of the disease, not considering the possible impact of the previous pharmacological treatment on platelet function, and different laboratory methodology as the use of dissimilar ways of labeling the receptor (Table 1).

Table 1.

Studies Focused on Platelet 5-HT2A Function

Abbreviations: BD, bipolar disorder; MDD, major depressive disorder; Dep, indicates a diagnosis of major or minor depressive episode, so it can be included in the context of a BD, a unipolar depression or a Schizoaffective disorder; UD, unipolar depression

Through the analysis of these studies, it is not possible to draw definitive conclusions. The results are controversial and sometimes conflicting. However, it is possible to identify some trends: within the whole sample of depressed patients, women, and subjects with suicidal behaviors would seem to have slightly higher B _max values at baseline, and the treatment with TCAs would seem to cause a greater increase in B _max values compared to other ADs.

In any case, it should be highlighted that the majority of these studies was carried out some decades ago and the ensuing findings would require to be replicated in larger samples and within the frame of the latest and more comprehensive pathophysiological models of depression. With no doubt, the platelet 5-HT2A receptors in depressed subjects appear to be overexpressed and to have a greater sensitivity and reactivity. This would imply a hyperresponsivity to platelet activation factors. It is possible to hypothesize that neuroendocrine and neuroinflammatory alterations may affect platelet reactivity in subjects suffering from depression.

Platelet 5-HT reuptake transporter (SERT)

Serotonin transporter, SERT (SLC64A for solute carrier family 6, member A4), is a 12 transmembrane domain (TMDs) protein, containing two sites of N-linked glycosylation. It performs the uptake of 5-HT through dissipation of the Na⁺ gradient established by the electrogenic pump Na/K ATPase. Serotonin is transported from the synaptic cleft back to presynaptic neuron where it terminates its action. This transporter is mainly located in cholesterol-rich membrane microdomains, also called lipid-rafts that act as platforms for the regulated assembly and functioning of signaling receptors and transporters.Reference Allen, Halverson-Tamboli and Rasenick ¹⁵⁴ The N- and C-terminal regions of SERT dip into the cytosol and interact with several proteins. Cytoplasmic domains, located between TMDs, also contain sites of posttranslational modifications. Regulating extracellular 5-HT, SERT influences the magnitude and duration of the 5-HT response. The SERT encoding gene was first cloned from rat brain and basophilic leukemia cells in 1991,Reference Blakely ^{155 ,} Reference Hoffman, Mezey and Brownstein ¹⁵⁶ while the human SERT gene was cloned 2 years later; it is located on chromosome 17q11.2 and contains 14/15 exons spanning around 40 kb.Reference Ramamoorthy ¹⁵⁷

Due to limitations in access to study the function of brain SERT directly in humans, it has been difficult to make inferences about the activity of neuronal SERT in clinical studies. Platelet membranes contain SERT and are easily accessible, so that they have been and are widely used to explore SERT that is one of the major biological parameters studied in neuropsychiatry. Although several studies suggested a high genetic and functional affinity, it is still whether platelet SERT fully correlates with brain SERT.Reference Pletscher and Laubscher ^{85 , 157–165} The kinetic characteristics of the SERT have been investigated throughout the decades by means of [³H]-imipramine ([³H]-IMI), [³H]-paroxetine ([³H]-Par), or [³H]-citalopram binding, in parallel with development of SSRIs, or through [³H]-5-HT and [¹⁴C]-5-HT reuptake parameters. Only a few studies used other specific ligands to test the pharmacokinetic characteristics of the platelet SERT because the data are meager and they did not comply with the selection criteria, they will be not reviewed in this paper.

Platelet 5-HT concentration

Ten controversial longitudinal studies and six cross sectional studies were selected.

In a longitudinal study, platelet 5-HT levels were measured by liquid chromatography with electrochemical detection (LCEC) in MDD, dysthymic disorder, schizophrenia, and other psychosis patients. A significant reduction in platelet 5-HT levels was found in MDD patients and in female schizophrenic patients, with no significant differences between drug-free or medicated patients.Reference Le Quan-Bui ¹⁹⁶ Similarly, two different clinical trials, using spectrofluorimetric method, reported significantly lower basal platelet 5-HT in drug-free, depressed patients than healthy subjects. In one of them, a treatment with amitriptyline or paroxetine provoked a further decrease in platelet 5-HT content.Reference Schlake ¹⁹⁰ These findings are affected by a main bias, as the subjects were not matched for sex and age, with the healthy group composed generally by young subjects (mean age: 27.5 years) with a discrepancy of about 14 years, compared to the depressed group. In the second study, platelet 5-HT was determined after 4 and 24 weeks of sertraline treatment in 15 female non-suicidal, nonpsychotic MDD patients, compared with 15 drug-free healthy women. Depressed patients were subdivided according to the treatment response into remitters, responders, and nonresponders based on, respectively, the 70%, 50 to 69%, and <49% reductions in MADRS scores. Despite the small number of patients, sertraline induced a further decrease in platelet 5-HT concentration, while suggesting that pretreatment values of platelet 5-HT did not predict therapeutic outcome to sertraline treatment at least in female depressed patients.Reference Pivac ¹⁹⁷ Another longitudinal study detected a baseline decrease of platelet 5-HT, as assessed by the HPLC, and showed a further decrease after both clomipramine and fluoxetine.Reference Alvarez ¹⁴¹ Similar results were obtained in an 8-week clinical trial with paroxetine carried out in 11 patients with postpartum depression, with platelet 5-HT determined with both immunocytochemical assay and HPLC. Although 5-HT levels were significantly lower in responders, compared with nonresponders, the small sample size makes this finding a mere observation.Reference Maurer-Spurej, Pittendreigh and Misri ¹⁹⁸ On the other hand, four longitudinal studies reported no baseline intergroup differences. In one study, the mean platelet 5-HT level resulted normal in 30 depressed patients compared with 20 healthy subjects, and decreased after 2 to 4 weeks of antidepressant treatments (TCAs, SSRIs, and MAOIs), independently from the clinical outcome.Reference Karege ¹⁹⁹ The same result was obtained in a 6 weeks open-label study with sertraline in 21 MDD patients,Reference Markovitz ¹¹⁷ or in 36 female MDD patients after 4 weeks of treatment with paroxetine or tianeptine (in this case, 5-HT content was assessed by spectrofluorimetric method).Reference Muck-Seler ²⁰⁰ As previously mentioned, the study by Zhuang et alReference Zhuang ¹⁹⁴ also reported no significant baseline differences between the depressed and the control group, while using the enzyme-linked immunosorbent assay to determine the platelet concentration of 5-HT. Again, SSRI administration for 4 weeks significantly reduced platelet 5-HT content.

Taking cross-sectional studies into consideration, five reported no statistically significant differences between the depressed and the control subjects, with three using the same HPLC despite, but including heterogeneous samples.Reference Roggenbach ^{148 ,} Reference Franke ^{191 ,} Reference Uebelhack ¹⁹³ One study, although underlying no difference between patients and controls, noted that higher platelet 5-HT concentrations were more evident in psychotic than in nonpsychotic patients. Furthermore, the lowest platelet 5-HT concentration was associated with the most severe self-aggressive behaviors (both suicidal attempts and completed suicides).Reference Muck-Seler ^{201 ,} Reference Muck-Seler, Jakovljevic and Pivac ²⁰² An exploratory study examined associations between cortisol parameters, platelet 5-HT content, and platelet activity markers (CD40, CD40L, and CD62P) in 19 patients with type 2 diabetes (T2DM) and untreated comorbid MDD, 24 T2DM patients, 21 patients with untreated MDD, and 25 age- and sex-matched healthy control subjects, by using HPLC with UV detection. Subgroups did not differ in 5-HT or cortisol slope, but in patients with both T2DM and MD, 5-HT and CD62P were positively correlated.Reference Zahn ²⁰³ In any case, this study suffers from some limitations: the cross-sectional design, the relatively small sample size, so only medium to large effects could be detected with sufficient statistical power; the cortisol samples were measured on a single day, so they might reflect an actual state rather than chronic burden; the exclusion of patients with cardio- and cerebrovascular disease, or peripheral artery disease, might only refer to T2DM patients without macrovascular diseases.

Finally, one cross-sectional study detected reduced platelet 5-HT content. This study examined the relationship between platelet 5-HT, plasma cortisol, and PRL concentrations in 20 schizophrenic, 25 depressed, and 25 healthy women. At the time of blood sampling, the schizophrenic and depressed patients had been drug-free for at least 7 days. Platelet 5-HT, plasma cortisol, and PRL concentrations were determined by spectrofluorimetric, radioimmunoassay, and immunoradiometric methods, respectively. Platelet 5-HT concentration was higher in schizophrenic patients than in depressed patients or in healthy control subjects, while it was lower in depressed patients than in healthy subjects or in schizophrenic patients. Plasma cortisol levels were significantly increased in both schizophrenic and depressed patients compared with values in healthy controls. The values of plasma PRL were similar across groups. Unfortunately, this study did not determine any relationships between platelet 5-HT concentrations and plasma PRL and/or cortisol levels after a neuroendocrine challenge test, and reported no information on phases of the menstrual cycleReference Muck-Seler ²⁰⁴ (Table 5).

Table 5.

Studies on Platelet 5-HT Concentration

Abbreviations: Dep, depressive episode so it can be included in bipolar disorder, unipolar depression, and Schizoaffective disorder; MDD, major depressive disorder; UD, unipolar depression.

Taken together, the impact of studies on platelet 5-HT contents is limited, given the main contribution of peripheral 5-HT and no clear evidence of any relationship with CNS 5-HT. Not surprisingly, the literature on this topic dates back to some decades ago, with a few exceptions.

Adrenergic system

A hypothetical dysfunction of the adrenergic system has been discussed for a long time as a cause or predisposing factor to the onset of depressive disorders.Reference Cottingham and Wang ^{206 ,} Reference Stahl ²⁰⁷ Considering that platelets possess adrenergic receptors similar to those present on neurons, alterations of the adrenergic pathway has been also examined in the platelets from depressed subjects. Adrenaline, through activation of α ₂-adrenergic receptors, provokes direct effects on platelets: it induces primary aggregation, inhibition of adenylate cyclase, potentiation of stimulus-induced aggregation, and secretion.Reference Siess and Lapetina ²⁰⁸

Investigations of α ₂-adrenergic receptors B _max of depressed patients led to inconsistent findings, as some studies reported heterogeneous results.

In the present review, we selected 13 longitudinal studies and 4 cross-sectional studies, investigating the receptor affinity parameters of the α ₂-adrenergic receptor. In addition, we included, in this section, four studies on platelet aggregatory response after adrenaline and noradrenaline and one evaluating intracellular calcium sensibilization with adrenaline.

Although different ligands (ie, [³H]-clonidine, [³H]-yohimbine, [³H]-rauwolscine, [³H]-para-aminoclonidine, [³H]-adrenaline, p¹²⁵I-clonidine, [³H]-UK14304) were used at baseline, 10 out of the 13 clinical trials reported increased receptor density, one decreased, and two reported no difference between depressed and healthy control groups. Two studies detected differences in baseline K _d. All four longitudinal studies that used [³H]-clonidine reported an increased baseline B _max in patients. ^209–212

Moreover, in one of these studies, the administration of either imipramine or amitriptyline led to significant decreases in the B _max and in the K _d.Reference Garcia-Sevilla ²¹⁰ The authors were the first to hypothesize that depression might be related to a α ₂-adrenergic receptor supersensitivity, and that the clinical effectiveness of TCAs was associated with a decrease in their number.Reference Garcia-Sevilla ²¹⁰ Along the same hypothesis, in another study, the binding of [³H]-clonidine hydrochloride to platelet membranes was measured in 13 MDD patients and correlated with the aggregation response induced by adrenaline hydrochloride, which is the result of the activation of the “high-affinity state” receptor. The B _max and K _d, as well as the aggregation response, were increased in the patients, and were negatively correlated. Long-term treatment (3-24 months) with lithium carbonate was associated with a statistically significant decrease in B _max and with an increase in the aggregation response. Moreover, the decreased number of platelets α ₂-adrenergic receptors was related to the duration of the lithium treatment. However, several limitations of this study should be underlined: the small size sample and the inclusion in the depressed group of patients with both unipolar and depression. Some patients were treated for only 3 months and others for 24 months, so that it is difficult to ascertain the long-term impact of lithium on morphology and platelet functions.Reference Garcia-Sevilla ²⁰⁹

The same authors partially confirmed these findings in two other clinical trials, using different patient samples and ligands. In the first case, they measured the specific binding of the full agonist [³H]-adrenaline in 14 drug-free depressed patients with melancholia, while reporting an increased B _max compared with 15 control subjects. Long-term administration of clomipramine was associated with decreased platelet α ₂-adrenergic receptor density correlating with the duration of treatment.Reference Garcia-Sevilla ²¹³ In the second one, using quantitative immunoblotting with anti-α ₂-adrenergic receptors, increased immunoreactivities were noted in 22 depressed patients, compared with 22 matched controls. They also showed that α ₂-adrenergic receptor/G_ai complex was associated to a decreased receptor kinase GRK-2. Moreover, a treatment with mirtazapine reversed this abnormality and induced downregulation of a₂-adrenoceptor/Gai complex. This suggests that a defect in the regulation of platelet GRK-2 might contribute to the upregulation of α2-adrenergic receptor.Reference Garcia-Sevilla ²¹⁴

A further support to this hypothesis was derived from another study, evaluating the kinetic parameters of [³H]-clonidine and [³H]-yohimbine in 29 MDD patients and 26 control subjects. At baseline, the authors reported an increase in B _max of [³H]-clonidine binding. Treatment with either imipramine hydrochloride or with amitriptyline hydrochloride decreased the B _max. Electroconvulsive therapy also decreased the specific binding of both [³H]-clonidine and [³H]-yohimbine, while lithium carbonate decreased the B _max of [³H]-yohimbine.Reference Smith ²¹¹ Another study, by using [³H]-yohimbine, reported higher B _max in 23 drug-free MDD patients than in 27 healthy control subjects. High pretreatment agonist affinity to the α ₂-adrenergic receptors predicted positive treatment outcome. Imipramine induced uncoupling (−11%) and redistribution of receptor density in treatment responders, with no effect on nonresponders.Reference Gurguis ²¹⁵

Another 4 to 6 week clinical trial compared [³H]-clonidine binding in platelets from 33 controls, 24 unipolar depressive, 22 schizophrenic, 18 bipolar patients, and 8 schizoaffective disorder during a drug-free period. The results showed that B _max in depressed, schizophrenic, and schizoaffective patients was significantly higher than in normal subjects. Desipramine treatment caused a significant increase in the K _d of [³H]-clonidine binding, with no changes in B _max.Reference Pandey ²¹² Two different studies of the same group,Reference Piletz ^{216 ,} Reference Piletz ²¹⁷ using [³H]-para-aminoclonidine and [p¹²⁵I]-clonidine, showed an upregulation of α ₂-adrenergic receptors in depressed patients returning to normal values after a 6- to 8-week treatment with desipramine.Reference Piletz ^{216 ,} Reference Piletz ²¹⁸ A single longitudinal study including 26 MDD inpatients and outpatients and 26 control subjects reported increased B _max and K _d values, by using [³H]-UK14304 as ligand. Chronic administration of mianserin to eight depressed patients slightly increased the B _max and K _d values.Reference Kaneko ²¹⁹ Clinical trials using [³H]-rauwolscine as ligand led to discordant results. The B _max and K _d were found similar in 19 hospitalized, untreated depressed patients vs 26 control volunteers. After 7 to 12 days treatment with different TCAs, a significant improvement was observed in the depressive symptoms, with no change in [³H]-rauwolscine binding parameters that also remained stable after 23 days of treatment, in spite of the clinical improvement, as shown by the decrease of the HRSD total score.Reference Pimoule ²²⁰ On the contrary, 29 unmedicated MDD patients showed decreased B _max values compared to 14 normal volunteers, with no differences in K _d value. The K _d values of the [³H]-rauwolscine binding were significantly higher in depressed patients treated with TCAs than in the unmedicated or treated with fluoxetine.Reference Maes ²²¹ A single study, involving 30 depressed patients and 30 healthy control subjects, with the use of [³H]-UK 14304, identified a decreased agonist-receptor affinity in depressed patients, whereas the B _max was not altered. After 3 months of antidepressant treatment, a decreased K _d (increased affinity) was observed, together with a decrease in B _max.Reference Karege ²²²

Even considering the cross-sectional studies, it is possible to identify a trend toward an increase in the α ₂-adrenergic receptor density or B _max. Specific binding of [³H]-clonidine was measured in 13 unipolar depressed elderly patients (mean ± SD age = 74.8 ± 5.9 years) and in 10 elderly controls (mean ± SD age = 75.2 ± 5.9 years). The B _max was significantly higher in depressed than in the controls, whereas K _d was not significantly different. In any case, these results are questionable, giving the small sample size and no consideration of the medical comorbidities or drugs that could have affected platelet function.Reference Doyle ²²³ Increased B _max values were also observed in 18 unipolar depressed patients versus 24 sex- and age-matched, healthy control subjects, using [³H]-para-aminoclonidine as ligand. This study, unlike many others, by using purified plasma membranes, ruled out the possibility that an inhibitor may have masked receptor binding in other studies, which used total platelet lysates.Reference Piletz ²¹⁷ On the other hand, a cross-sectional study, using [³H]-para-aminoclonidine as well to investigate the affinity binding, reported significantly lower B _max values in 26 MDD depressed than in 22 control subjects.Reference Carstens ²²⁴

Finally, one study, analyzing a sample of 15 unipolar depressed and 15 age-and sex-matched control subjects, showed no difference in [³H]-rauwolscine B _max and K _d between the groups, but a significant and positive correlation between B _max and HRSD total score.Reference Marazziti ²²⁵

Otherwise, it should be briefly pointed out that several authors measured adrenoceptor-mediated platelet primary aggregation in depressed patients and healthy subjects. Both initial velocity and maximum amplitudes of platelet response to increasing concentrations of adrenaline were decreased in drug-free depressed, while suggesting a postsynaptic desensitization of platelet adrenoceptor function. ^226–228 In a longitudinal clinical trial, the functional status of platelet α ₂-adrenergic receptors in 30 MDD patients was assessed by simultaneously measuring both a biochemical mechanism of transduction of receptor activation (inhibition of adenylate cyclase activity) and a physiologic response of the receptor (induction of aggregation). The inhibitory effects, induced by adrenaline and [³H]-UK-14304 on adenylate cyclase activity, were unchanged, while, contrary to the previously mentioned studies, the aggregation responses induced by the same α ₂-adrenergic receptor agonists were potentiated, which indicated receptor supersensitivity. Both functional responses are desensitized after long-term antidepressant treatment.Reference Garcia-Sevilla ²²⁹ A single cross sectional study, analyzing a sample of 20 MDD patients and 20 control subjects, revealed no significant intergroup differences in adrenaline and noradrenaline aggregatory responses.Reference McAdams and Leonard ²³⁰ Finally, Mikuni et alReference Mikuni ²³¹ used noradrenaline and 5-HT to assess intracellular calcium mobilization in blood platelets from 11 depressed patients and 11 healthy control subjects. The 5-HT response resulted in a significant greater mobilization of intracellular calcium in depressed subjects than in control subjects, while noradrenaline provoked a higher, albeit not statistical intracellular calcium influx in the platelets of depressed patients (Table 6).

Table 6.

Studies on α2-Adrenergic Receptor Function

Abbreviations: Dep, depressive episode so it can be included in bipolar disorder, unipolar depression, and Schizoaffective disorder; MDD, major depressive disorder; UD, unipolar depression.

Most of the studies suggest that a noradrenergic receptor responsiveness may be abnormal in the affective disorders and may be altered by long-term antidepressant treatment. Several authors speculated on the possible dysfunctional noradrenergic states that may differ between patient subgroups. Specifically, depressed subjects would tend to denote a “hyper-reactive” adrenergic system, not only compared to healthy control subjects, but also to other psychiatric disorders. Unipolar depressed patients should be characterized by a condition of α₂-adrenergic postsynaptic subsensitivity accompanied by episodic increase of central presynaptic noradrenergic availability. If the increased number of platelet α₂-adrenergic receptors in patients with depressive disorder is relevant to the disease and reflects a state of supersensitivity of similar receptors in the brain, it follows that effective antidepressant treatment should induce receptor changes in the opposite direction. Different antidepressant treatments (SSRIs, TCIs, ECT, lithium salts, and mirtazapine) reduce the B _max of high-affinity binding sites for many ligands ([³H]-clonidine, [³H]-yohimbine, [³H]-adrenaline, [³H]-para-aminoclonidine, [³H]-UK14304, and p¹²⁵I-clonidine) on platelet membranes. However, this decrease in the number of α-adrenergic receptors was not associated with an increased affinity.

Nevertheless, due to the discrepancies attributable to several methodological variables, such as heterogeneous platelet or sample sizes and the choice of radioactive ligand and competitor (nonradioactive ligand) in the assay, it is not possible to draw definitive conclusions.

The strikingly greater variances in patients’ B _max and K _d values observed in these studies would be more consistent with the hypothesis that depressed patients might have an abnormal reactivity of the adrenergic system.

In any way, the role of the noradrenergic system in depression seems to be ancillary to that of other neurotransmitters, systems, and processes, and the impact of platelet studies’ results was limited and poorly explored in the last decades.

Platelet mono-amine oxidase activity

MAO is a mitochondrial outer membrane flavoenzyme catalyzing the oxidative deamination of neurotransmitters and exogenous arylalkylamines, widely expressed throughout central and peripheral tissues. Two MAO isoforms exist in mammals, namely MAO-A and MAO-B, which are products of distinct genes both located on the short arm of the X chromosome. MAO-A and MAO-B genes comprise 15 exons with identical intron-exon organization. Substrates for MAO include monoaminergic neurotransmitters, such as catecholamines (ie, dopamine, norepinephrine, and epinephrine) and 5-hydroxy-tryptamine (serotonin). MAO isoenzymes display different substrate specificity. However, in living organisms, the contribution of MAO-A and MAO-B in the deamination of substrates depends not only on their kinetic properties, but also on the relative distribution and abundance in the different organs and brain regions. Both MAOs are more distributed in neurons and in the astrocyte, but the presence outside the CNS is more differentiated: while MAO-A are also present in the liver, pancreas, spleen, gastrointestinal tract, and placenta, MAO-B predominates in skin and skeletal muscle, and is the sole form in platelets. Human brain MAO is 70 to 80% of type B.Reference Blaschko ^{149 ,} Reference Yeung ²³² Platelets also possess MAO-B, generally considered an unspecific biological vulnerability marker toward some neuropsychiatric conditions (such as mood, psychotic eating, post-traumatic stress disorder, etc.), or related to personality traits. ^233–243

Using the selection criteria indicated above, we included four longitudinal studies and one cross-sectional study in this review.

In one of the first studies, the platelet MAO activity was investigated in unipolar and bipolar depressed patients with the substrates ¹⁴C-tyramine and ¹⁴C-benzylamine. Sixty-two patients were selected from consecutive admissions over a period of 2 years. They were untreated with psychotropic drugs (other than benzodiazepines) for at least 2 weeks, or by ECT, when necessary. A group of 59 healthy subjects and 20 in-patients with neuroses and mild depression were chosen as control groups. A higher mean platelet MAO activity, assessed by the substrate tyramine, but not benzylamine, was found in depressed than in control subjects that also resulted positively associated with illness severity. In any case, the increase was mainly due to bipolar patients. The ECT provoked no changes upon the platelet MAO activity of 15 depressed patients. On the contrary, a lithium treatment for approximately 6 months produced a statistically significant rise in platelet MAO activity with the substrate benzylamine but not tyramine. In the opinion of the author, the differences in platelet MAO activity in depressed patients compared with controls are due to MAO structural differences. The form of the enzyme may be the mechanism through which genetic factors influence both the type and severity of the affective illnesses. Furthermore, the difference in MAO activity between unipolar and bipolar patients could suggest a genetic effect acting as a predisposing factor to illness awaiting an appropriate stress.Reference Mann ²⁴⁴

In support of this hypothesis, another longitudinal study, assessing the platelet MAO activity in 25 MDD patients and 25 depressed patients (unmedicated at baseline and after 3 weeks and 2 months of imipramine) using 5-HT as substrate, showed a higher activity of the enzyme that progressively decreased after the treatment and in parallel with the clinical improvement.Reference Quintana ²⁴⁵ In contrast, other studies reported no significant differences between depressed and control subjects. In one open study, no difference was found between 36 female depressed patients (21 treated with tianeptine and 15 treated with paroxetine) and 11 drug-free healthy women, using kynuramine as substrate, before and after treatment. Furthermore, there were no correlations between MAO activity, peripheral biomarkers, platelet and serum 5-HT concentrations, plasma levels of cortisol and prolactin (PRL) either before or after antidepressant treatment.Reference Muck-Seler ²⁰⁰ The MAO activity measured by spectrofluorometer at baseline was similar in 15 female non-suicidal, nonpsychotic MDD patients and 15 drug-free healthy women. Long-term sertraline treatment induced a statistically significant decrease in platelet MAO activity after 24 weeks, in parallel with symptom remission.Reference Pivac ¹⁹⁷ Finally, another cross-sectional study detected no difference in platelet MAO activity, with [¹⁴C]-phenylethylamine as substrate, among 60 suicidal inpatients, 28 non-suicidal depressed inpatients, and healthy control subjects. A lower MAO-B activity was observed only in suicidal female patients. In this case, the authors used [¹⁴C]-phenylethylamine as substrate.

Several explanations might be put forward to this inconsistency, specifically, the use of heterogeneous biochemical methods and/or substrates, different diagnostic systems, or sample characteristics and eventual treatments that are all important sources of variation. Other sources of variation of the MAO-B activity are sex, age, medications, smoking habits, alcohol, or illicit drugs (Table 7).

Table 7.

Studies on Platelet MAO-Activity

Abbreviations: Dep, depressive episode so it can be included in bipolar disorder, unipolar depression, and Schizoaffective disorder; MDD, major depressive disorder; UD, unipolar depression.

In any case, the studies on platelet MAOs, although eliciting a great interest in the past decades, have progressively declined, especially in the field of pathophysiology of depression.

Biomarkers of platelet activation

As already mentioned in the introduction, platelets are now recognized as key players in inflammatory and innate immune responses, for their ability to interact with almost all immune cells. ^52–62 Understanding platelet activation pathways and potential biomarkers might represent important clinical targets for several diseases, as well as new diagnostic and therapeutic possibilities in monitoring the disease activities and responses to treatment. Upon activation, platelet surface P-selectin is overexpressed, and platelets release their granule contents into circulation. Several markers of platelet activation, such as P-selectin, CD-40, CD-40L, sCD-40L, CD-63, β-thromboglobulin, PF4, and GPIIb/IIIa, have been identified to correlate with the presence of inflammation and atherosclerosis. ^246–249 Platelet activation markers can be well studied by ELISA or western blot, but flow cytometer is currently the standardized method to study platelet function. ^{250 ,} Reference Lacroix ²⁵¹

One of the first studies investigated the degree of platelet activation in a sample of 15 MDD patients and 12 age-matched healthy subjects.Reference Musselman, Marzec and Manatunga ²⁵² At baseline, the depressed subjects exhibited greater procoagulant activity than controls, as detected by increased binding of the monoclonal antibodies anti-ligand-induced binding site (mAb anti-LIBS), which attaches to the fibrinogen binding site of the activated GP receptor, increased binding of the mAb GA6 to P-selectin, and increased plasma concentrations of the platelet-specific secretion protein PF4. The increased binding of the anti-LIBS reflects the conformational change in GP IIb/IIIa. After 6 weeks of open-label treatment with paroxetine, the depressed patients exhibited significant reductions in all three parameters with a normalization of the platelet activation. Although interesting, this study was carried out in a small sample with a lack of normal comparison subjects with other risk factors for ischemic heart disease (IHD). Given that the increased risk of IHD or cerebrovascular disease in MDD subjects might be due to an increase in platelet activity, on the basis of these findings, the authors speculated that a medium or long-term antidepressant treatment, by normalizing the degree of platelet activity, might reduce the cardio-cerebrovascular risk. Subsequently, the same authors provided further supporting evidence to their hypothesis. Markers of in vivo platelet stimulation and secretion were measured in 12 normal subjects and three patient groups: 15 MDD patients, 12 dialysis-dependent patients, and 10 patients with severe thoracic aortic atherosclerosis (TAA), transesophageal echocardiography (TEE)-documented. In comparison with the control subjects, depressed patients and patients with thoracic aortic atherosclerosis exhibited the greatest platelet stimulation, as detected by increased mAb anti-LIBS platelet binding. These findings extend previous observations of increased platelet activation not only in MDD patients but also in those with thoracic aortic atherosclerosis.Reference Musselman ²⁵³

Other authors used western blotting and revealed more P-selectin immunoreactivity in depressed patients with no cardiovascular history or other medical illness, compared to healthy individuals. Patients were assessed after 6 to 8 weeks of bupropion treatment, and it emerged that P-selectin remained high in depressed patient and was not related to HRSD score or with the post-treatment plasma concentrations of bupropion. In the opinion of the authors, the elevation of P-selectin could be considered a trait marker for depression.Reference Piletz ¹¹⁶

Supports to the presence of platelet hyperactivity in depression derived from a study reporting an increased expression of GPIb receptors, CD-62, and P-selectin, while Gpα_IIB/β_IIIa and CD63, although also increased, did not reach the significant difference (MDD = 15; controls = 15). There was no difference between depressed patients and healthy volunteers in the expression in glycocalicin or vWF plasma concentration, or ADP-induced aggregation. Of note, the samples were not perfectly matched by age. The results of this study demonstrated that the number of GPIb receptors on platelets are increased in depression and would perhaps suggest a novel risk factor for thrombosis in MDD patients.Reference Walsh ²⁵⁴ In addition, in the same study, smoking has also been shown to increase CD62 on platelets, reflected significantly higher expression of CD62 observed among smokers versus nonsmokers in the control group. However, the causal relationship between smoking and depression is yet to be fully elucidated.Reference Walsh ²⁵⁴

More recently, another study, using whole blood aggregometry, flow cytometry, and ELISA, investigated the platelet expression of P-selectin, CD40-CD40L (atherosclerotic marker), PF-4, and β-thromboglobulin levels in 21 depressed patients and 25 healthy age- and sex-matched controls. All subjects were taking no antiplatelet medications within 7 days before sampling, and had no clinical evidence of peripheral artery disease, coronary or cerebral atherosclerotic, or inflammatory disease based on clinical history and examination. They also had no significant atherosclerotic risk profile except smoking. The results showed that P-selectin, CD40, CD-40 L, sCD40L, and β-thromboglobulin were significantly higher in depressed than in the control subjects, while CD41, CD42b, and PF-4, as well as the aggregometry results, did not differ between the two groups. The increase of these platelet markers in newly diagnosed patients would emphasize that depression is linked to a prothrombotic and proinflammatory state, possibly contributing to accelerate atherosclerosis.Reference Neubauer ²⁵⁵

One very interesting exploratory study examined cortisol awakening response (CAR), diurnal decrease in salivary cortisol concentrations (slope), and the expression of CD40, CD40L, sCD40L, p-selectin, β-thromboglobulin, and PF-4 in 20 MDD patients, and compared to 18 MDD patients with type II diabetes (T2DM) with MDD, and 24 healthy controls. Subgroups did not differ in 5-HT or cortisol slope, while T2DM patients without MDD showed significantly lower CAR than did healthy control subjects. Platelet markers were elevated in depressed and diabetics, while P-selectin and 5-HT levels were positively correlated in depressed patients only. The measurement of plasma cortisol on a single day and the relatively small sample represent the major limitations of this study; again, the cortisol parameters might reflect an actual state rather than chronic cortisol burden.Reference Zahn ²⁰³

A pilot study evaluated platelet 5-HT levels and biomarkers of platelet activation in platelets of untreated (n = 13) and SSRI-treated (14) depressed patients and normal subjects (n = 14). Depressed patients showed a significant reduction in platelet 5-HT, with no change after treatment. The expression of P-selectin, β-thromboglobulin, CD-63, and Gpα_IIB/β_III was similar in the different groups. It should be emphasized that the author did not provide information on the diagnostic assessment and did not show the values of the aforementioned proteinsReference Maurer-Spurej, Pittendreigh and Solomons ²⁵⁶ (Table 8).

Table 8.

Studies on Biomarkers of Platelet Activation

Abbreviations: Dd, dialysis-dependent patients; Dep, depressive episode so it can be included in bipolar disorder, unipolar depression, and Schizoaffective disorder; MDD, major depressive disorder; TAA, thoracic aortic atherosclerosis patients; UD, unipolar depression.

Most of the results emerging from the analyzed studies would seem to suggest that depressed subjects have an abnormal platelet activity. Based on the specific molecules and ligands secreted and expressed, platelets would seem to present a hyperactivation state. In other words, a depressive episode or a depressive disorder, through unidentified (neuro)-inflammatory mechanisms, would result in a pro-inflammatory and pro-thrombotic state. This condition, enhancing atherosclerosis processes, could possibly explain why depressed patients are more prone to develop IHD. This hypothesis would imply that an effective antidepressant treatment, managing the symptoms presented by the patients, could reduce the pro-inflammatory state, reducing the risk for IHD. At the same time, patients with a history of depressive disorder should seriously consider the introduction of drugs that reduce platelet activity, lowering the risk for IHD. Although there are still a few studies, SSRIs would appear the safest and the most efficient antidepressants to reduce platelet activation status in depression, although a debate is still open on their potentially increased risk for major bleeding (especially in combination with platelet inhibitor drugs or oral anticoagulants). ^258–264

Platelet hyper-responsivity

Platelet hyper-reactivity is not merely an unavoidable adverse effect of conditions ranging from infections to diabetes, it is a strong factor exacerbating most aging-related human pathologies.Reference Podrez ²⁶⁵ Taken together, recent studies indicate that platelet reactivity is not just a biomarker of adverse effects, but it is by itself a condition that requires management, not limited to cardiovascular disease and stroke. The mechanisms, leading to platelet hyper-reactivity, include variations in platelet proteins, such as surface receptors or changes in the balance between procoagulant and anticoagulant factors in plasma. Several authors investigated platelet responsiveness to pro-activating agents, such as ADP, 5-HT, adrenaline, collagen, and thrombin in psychiatric patients, mainly in depressed patients, with somewhat controversial results. Some of them reported greater platelet reactivity with altered mobilization of intracellular calcium (Ca²⁺) deposits in response to stimulation with the aforementioned agents. In several studies, 5-HT stimulated Ca²⁺ mobilization was found increased in depressed patients, compared with the controls.Reference Mikuni ^{231 , 266–273} In one of these studies, the 5-HT stimulated Ca²⁺ response was significantly higher in the platelets of unmedicated depressive patients, with no intergroup difference in either basal Ca²⁺ concentration or thrombin-induced Ca²⁺ mobilization. In two different studies, Kusumi et alReference Kusumi, Koyama and Yamashita ²⁶⁷ found a significantly higher 5-HT-stimulated response in unmedicated patients with bipolar depression and melancholic MDD than in those with non-melancholic MDD and normal controls. The enhanced Ca²⁺ mobilization failed to correlate with severity of depressive symptoms. In patients with BD and melancholic MDD, a significant difference in 5 HT-stimulated Ca²⁺-response was noted between the unmedicated group and those in remission.Reference Kusumi, Koyama and Yamashita ^{266 ,} Reference Kusumi, Koyama and Yamashita ²⁶⁷ Interesting findings emerged from another study, in which not only there was a greater 5-HT response in 24 MDD patients than in 20 controls subjects, but, moreover, depressed patients with high anxiety showed significantly increased 5-HT responses compared to depressed patients with low anxiety. In addition, patients receiving SSRIs demonstrated lower 5-HT Ca²⁺ mobilization than those not taking drugs.Reference Delisi ²⁷⁰ One cross-sectional study investigated the intracellular free Ca²⁺ concentration in platelets of both untreated depressed patients and patients in remission treated with imipramine. The 5-HT-induced Ca²⁺ mobilization was significantly higher in both untreated patients and imipramine responders, compared with healthy controls. In addition, the responders showed a positive correlation between the duration of the remission and intracellular basal-free Ca²⁺ concentration, while suggesting that 5-HT-induced Ca²⁺ responses persisted after remission in depressed patients, and may perhaps be considered a trait marker.Reference Tomiyoshi ²⁷²

The increased mobilization of Ca²⁺ in depressed subjects resulted as a selective response to 5-HT, as other platelet activators, such as adrenaline,Reference Mikuni ²³¹ thrombinReference Eckert ²⁶⁸ or ADP,Reference Shimbo ²⁷¹ did not show an increased response.

Nevertheless, other authors described increased platelet secretion in response to some stimulants. One study detected a greater platelet secretion in depressed patients than in controls in response to collagen. Furthermore, patients with a family history of coronary disease had not significantly greater wound-induced fibrinogen receptor binding than the other groups. After treatment with sertraline, this higher responsiveness decreased.Reference Markovitz ¹¹⁷ Again, MDD patients showed a significantly higher platelet activation response to thrombin and/or collagen responsivity than patients with subsyndromal depression, schizophrenia, or controls.Reference Berk and Plein ^{274 ,} Reference Lederbogen ²⁷⁵ Antidepressants (SSRIs and TCIs) and ECT did not significantly change this apparent super-sensitivity.Reference Lederbogen ^{275 ,} Reference Pandey ²⁷⁶

One study examined the platelet glutamate receptor sensitivity using the platelet intracellular Ca²⁺ response to glutamate measured by spectrofluorometer. The results showed that unipolar depressed out- and inpatients showed a significantly greater platelet intracellular Ca²⁺ response to glutamate stimulation than the controls, in terms of both absolute values and percentage of response from baseline). In addition, depressed patients showed higher basal intracellular Ca²⁺ levels than controls, so that it is unclear whether the increased responsiveness to glutamate depends on a primary receptor super-sensitivity, or it is secondary to an alteration of the pathways of second messengers between the receptor activation and the mobilization of intracellular Ca²⁺ deposits.Reference Berk, Plein and Ferreira ²⁷⁷

Other authors detected no difference in platelet responsiveness to activating factors and matched controls, while showing an enhanced response to thrombin and PAF in bipolar.Reference Alvarez ^{141 ,} Reference McAdams and Leonard ^{230 ,} Reference Walsh ^{254 ,} Reference Maurer-Spurej, Pittendreigh and Solomons ^{256 , 278–282} Dubovsky et alReference Dubovsky ²⁸² did not find any differences in basal intracellular Ca²⁺ between unipolar depressed, BD patients, and healthy controls. Bipolar patients showed an enhanced response to thrombin and PAF, compared to unipolar depressed and control subjects. In another study, intracellular Ca²⁺ was measured in the resting state and, after stimulation with thrombin, platelet-activating factor, and 5-HT in 27 unipolar depressed, 17 BD patients, and their matched control individuals. No intergroup differences were found in the resting state and after stimulation, with the exception of intracellular Ca²⁺ after stimulation with 5-HT that was significantly higher in the lithium-treated group, whereas antidepressants and antipsychotics did not seem to interfere with it.Reference Dubovsky ²⁸²

In two different studies, no baseline differences were found in depressed patients and healthy controls in the percentage of 5-HT-amplified platelet aggregation to AD.Reference Gomez-Gil ^{279 ,} Reference Gomez-Gil ²⁸⁰ The aggregatory response significantly decreased with remission after imipramine treatment, perhaps through a desensitization or downregulation of the 5-HT_2A receptor.Reference Gomez-Gil ²⁷⁹ A 6-weeks clinical trial showed that SSRIs did not affect platelet aggregation and dense granule release in response to thrombin, but they significantly reduced ADP-induced platelet aggregation and dense granule release in both patient and normal control samples. According to the authors’ opinion, the active inhibition of platelet aggregation by SSRIs might explain their cardiovascular benefit.Reference Maurer-Spurej, Pittendreigh and Solomons ²⁵⁶

Finally, only one study identified a reduced platelet responsiveness to multiple activating factors (ADP, adrenaline, noradrenaline, 5-HT, collagen, thrombin, and ristocetin) in a small sample of 17 unipolar depressed patients, compared to 10 healthy controls, thus suggesting the presence of inhibitory factors in depressed patients, capable of reducing platelet aggregationReference Nugent, Dinan and Leonard ²²⁸ (Table 9).

Table 9.

Studies on Platelet Responsivity

Abbreviations: Dep, depressive episode so it can be included in bipolar disorder, unipolar depression, and Schizoaffective disorder; MDD, major depressive disorder; UD, unipolar depression.

All these findings suggest that the platelet response to activation in MDD patients is, in most of the case, supersensitive. This hyper-reactivity could lead to an increase in aggregability and consequently to an increased risk of thrombosis, representing a further confirmation of the close link between depression and cardiovascular disease, especially in untreated subjects. The positive effects of antidepressants on these parameters are noteworthy and deserve to be more thoroughly investigated in further studies.

Morphological changes

Five studies investigated eventual platelet morphological changes in depressed patients.

A first cross-sectional study conducted on a sample of 12 drug-free MDD patients compared to 22 healthy controls, reported a correlation between depression and discoid platelet form, and between HRSD score and platelet count.Reference Palmar, Marcano and Castejon ²⁸³ Furthermore, the patients showed dilated tubular and canalicular structures and an irregular and abundant glycocalyx on the outer membrane. The granules showed no changes in shape and size, although a denser structure was observed in the patients who also presented a greater amount of glycogen, or alterations of the mitochondria, which were smaller and more dilated than in healthy subjects.

Two studies deal with changes concerning the mean platelet volume (MPV). One of them reported elevated MPV in MDD patients (n = 15), as compared with controls (n = 17). Moreover, normalization of MPV levels and reduction of platelet count were observed after 8 weeks of treatment with escitalopram. However, this study is affected by the bias that the sample was relatively small and mainly composed of women in both patients (73.3%) and controls (64.7%). Moreover, the lack of control group with patients with other mental disorders or of patients treated with other types of antidepressant treatments (pharmacological and non-pharmacological) suggests the need of larger prospective trials to establish the clinical relevance of the described laboratory data.Reference Ataoglu and Canan ²⁸⁴

A cross-sectional study was carried out on a large sample of depressed patients (n = 287) compared with a control group of healthy subjects (n = 1999). The authors confirmed the increased MPV in depressed patients that was more pronounced among women. A significant negative correlation between MPV and platelet count was also found. However, despite the large sample size, the study suffers from some limitations: the preponderance of women in the whole sample, the lack of randomization of the selected subjects, the use of a list of questions following DSM-IV criteria, rather than a structured interview to make diagnosis, the lack of a distinction between bipolar and unipolar depressed patients, no assessment of illness severity and, finally, no consideration of other psychiatric or medical conditions that could potentially alter platelet functioning.Reference Canan ²⁸⁵

Two other studies detected no morphological changes of platelets in MDD. One of them investigated the effects on platelet shape induced by 5-HT in 29 MDD patients (12 treated with mianserin for 2-11 weeks) and 26 healthy controls, while reporting no intergroup difference. However, when the treated patients were compared with non-treated ones or healthy subjects, the first group showed an increased 5-HT-induced shape change. In the opinion of the authors, this would represent a sign of a marked influence on the intracellular signal transduction system due to the long-term treatment. However, it should be noted that the depressed patients were both bipolar and unipolar, were not sex-or age-matched with controls, and no information was provided regarding psychiatric and/or medical comorbidities that could potentially modify the results.Reference Kaneko ²⁸⁶

Finally, one study, by using phase-contrast microscopy, did not reveal any differences between unmedicated depressed and healthy controls in platelet morphology. These findings were confirmed by differential interference contrast micrographs. Platelets from control subjects and untreated depressed showed bright dots where the specific 5-HT antibody and the fluorescently labeled secondary antibody colocalize in dense granules. The platelets of the SSRI-treated patients showed differences in shape or in releasing dense granule content except for the lack of bright dots, while indicating very low levels of 5-HT in the dense granules. Generally, there was an inverse correlation between the SSRI dose and the number of 5-HT-labeled dense granulesReference Maurer-Spurej, Pittendreigh and Solomons ²⁵⁶ (Table 10).

Table 10.

Studies on Morphological Changes

Abbreviations: Dep, depressive episode so it can be included in bipolar disorder, unipolar depression, and Schizoaffective disorder; MDD, major depressive disorder; UD, unipolar depression.

In conclusion, the studies on morphological changes in platelets of depressed patients, although potentially interesting, are a few, heterogeneous, carried out with different techniques and, as such, not comparable.

Miscellany

Glycogen synthase kinase 3-beta (GSK3B) is a serine–threonine kinase with a key role in the regulation of the glycogen synthesis.Reference Embi, Rylatt and Cohen ²⁸⁷ GSK3B is also involved in the regulation of critical intracellular signaling pathways, including cell cycle, gene expression, and apoptosis.Reference Grimes and Jope ²⁸⁸ In neurons, GSK3B plays a major role in cytoskeleton organization and remodeling, synaptic plasticity, neurogenesis, and resilience to neuronal injury.Reference Bradley ^{289 ,} Reference Peineau ²⁹⁰ The studies highlighted the abnormal regulation of GSK3B in the pathophysiology of neuropsychiatric disorders and the impact of several psychotropic drugs (mood stabilizers, antidepressants, and atypical antipsychotics) on its activity. ^291–297

In the present review, we reported a cross-sectional study on GSK3B activity in a group of 40 drug-free MDD elderly patients and 20 age-matched controls. The GSK3B activity was indirectly inferred by the GSK3B ratio, pGSK3B/tGSK3B that was determined by enzyme immunometric assays. The lower GSK-3B ratio would have indicated a higher proportion of the non-phosphorylated GSK-3B and, thus, a higher enzymatic activity. As compared with age and sex-matched control subjects, the depressed patients showed higher GSK3B activity than the control group. By subdividing the depressed group basing on severity and the presence of cognitive impairment, it emerged that both severity and cognitive decline positively correlated to GSK3B activity. These findings suggested additional evidence of the involvement of GSK3B in the pathophysiology of late-life MD. Abnormal GSK3B activity may reflect a severe disruption of neuronal homeostasis and neurodegeneration with potentially important implications to distinguish psychiatric and neurodegenerative disorders. In any case, this study is affected by a major bias, which is the different numerosity of the patient and control groups. Given the impact of cognitive impairment on GSK3B activity, it is difficult to ascertain if the statistical difference between the two groups was mainly due to patients with cognitive impairment.Reference Diniz ²⁹⁸

Although these data highlight the possible role of intracellular mechanisms in MDD, they are just a few and should be considered as mere suggestions to be explored in the future.

Brain-derived neurotrophic factor (BDNF) is a growth factor protein and member of the neurotrophin family. BDNF contributes to a variety of neural processes ranging from neurodevelopment to the survival and homeostatic maintenance of central and peripheral nervous system in adulthood and in aging. ^299–301 During development, BDNF plays a key role in neurogenesis, differentiation, and maturation of neurotransmitter systems. ^302–304 In adulthood, it is agreed that BDNF plays an important role in the modulation of synaptic plasticity and in the formation of memory and learning. ^305–307 Animal models showed that BDNF is involved in the pathogenetic mechanism of MD, as well as postmortem studies showed an abnormal expression of BDNF in the hippocampus and prefrontal cortex of depressed patients. ^308–312 The morphological changes in the brain, reported in depressed patients, seem to be reverted by the long-term administration of ADs that would induce an upregulation of BDNF. ^313–316 BDNF is also present in large amounts in the blood where it is mostly stored in platelets (~99%), and only a small part of free BDNF is present in plasma. ^317–319 Furthermore, BDNF is present in two distinct pools in platelets, in α-granules, and in the cytoplasm.Reference Serra-Millas ³²⁰ BDNF in the α-granules is released upon platelet activation, whereas the cytoplasmic BDNF is not.Reference Tamura ³²¹ The maximum BDNF release is approximately 30% to 40% with stimulation, and the remaining 70% of BDNF is equivalent to that found in the cytoplasm, which is not released. ³²² Thus, the plasma level of BDNF only partially reflects the amount effectively produced by platelets.

BDNF levels in the human brain are difficult to measure, so it is generally assessed in periphery. The relevance of this measure is supported by animal studies, showing that BDNF is able to cross the blood–brain barrier in both directions, and that peripheral and central BDNF levels are associated. ^323–325 Since platelets cannot pass the blood–brain barrier, the BDNF level in the brain may not be reflected by the amount of BDNF associated with platelets, but rather by the amount of free BDNF in plasma.Reference Radka ³¹⁸ Drug-free depressed patients and healthy subjects with depressive personality traits show decreased serum BDNF levels, which would increase after antidepressant treatment. ^326–333 Some studies reported low BDNF levels in plasma of depressed subjects with gender differences, pointing out a specific role of gender in altering neurotrophin regulation.Reference Dwivedi ^{309 ,} Reference Karege ^{310 ,} Reference Gervasoni ^{329 , 334–340}

One cross-sectional study revealed that platelet BDNF levels were significantly lower in 20 suicidal depressed patients and 20 nonrecent suicidal depressed patients than in healthy controls. As suggested by the authors themselves, this study has some limitations: the sample size was small and the platelet BDNF levels were within a relatively wide range; the depressed group was heterogeneous, being established by 15 men and 45 women and, as already noted,Reference Huang, Lee and Liu ³³⁵ the reduction of serum BDNF levels in MDD might be typical to women, not men; finally, they analyzed only platelet-rich plasma and platelet-poor plasma BDNF levels. Therefore, it is difficult to determine if the source of BDNF decline is brain or other peripheral cells.Reference Lee and Kim ³⁴¹ Although several research studies described that changes in circulating BDNF may be used as a proxy for changes occurring in the brain,Reference Radka ^{318 ,} Reference Lommatzsch ^{342 ,} Reference Karege ³⁴³ it is uncertain how precisely BDNF levels measured in venous blood samples reflect circulating BDNF levels in vivo, and how different methodological approaches may affect the results.

To summarize, on the basis of the available information, the importance of assessing platelet BDNF appears limited both for the controversial results emerging from the studies and because there is no direct correspondence between BDNF produced by platelets and that released in plasma under stimulation.

Finally, two studies that analyzed endothelial nitric oxide synthase (eNOS) activity and plasma nitric oxide metabolites (NOx) levels in MDD patients were included in this review. Major depression has been associated with increased cardiovascular mortality in patients with coronary heart disease (CHD), and it has been described as an independent risk factor for the development of CHD in healthy subjects; however, the underlying mechanism remains unclear. It is known that nitric oxide plays a major role in cardiovascular regulation and its decreased production has been associated with several cardiovascular risk factors. In a cross-sectional study, Chrapko et alReference Chrapko ³⁴⁴ investigated the correlation between platelet endothelial nitric oxide synthase (eNOS) activity and plasma nitric oxide (NOx) levels. By comparing a sample of 15 depressed patients with healthy controls, NOx and eNOS resulted statistically significant decreased in the patient group. No correlation was found between the illness severity and eNOs activity or plasma NOx levels, or between platelet eNOS and NOx. In addition, by analyzing MDD patients with and without comorbid anxiety disorders, no difference was found. This would suggest a lack of a major confounding effect of comorbid anxiety disorder in these results, even if, due to the limited representation of each anxiety disorder in the sample, it cannot be excluded that certain anxiety disorders might be associated with specific dysregulation of endothelial NO production and platelet eNOS activity.Reference Chrapko ³⁴⁴

In a separate study with a different sample, the same research group confirmed the previous findings, and furthermore analyzed the effect of 8-weeks paroxetine treatment on NOx and eNOS in both depressed (n = 17) and controls (n = 12). After paroxetine treatment, plasma NOx levels increased in both samples. On the other hand, platelet eNOS activity was reduced in the control group, while no change was detected in MDD patients. According to the authors, this reduced production of NO by vascular endothelium and platelets could contribute to an increased risk of developing cardiovascular disease in depressed subjects. The impact of this risk factor could be reduced by the administration of paroxetine. It should be highlighted, however, that the small sample size and the choice of including in the depressed group people with other psychiatric disorders in comorbidity could be considered as limitsReference Chrapko ³⁴⁵ (Table 11).

Table 11.

Studies on Different Platelet Molecular Pathway

Abbreviations: Dep, depressive episode so it can be included in bipolar disorder, unipolar depression, and Schizoaffective disorder; MDD, major depressive disorder; UD, unipolar depression.

These studies are potentially interesting but deserve to be replicated in larger samples.