Dihydrotestosterone in Amyotrophic lateral sclerosis—The missing link?

Abstract Objective Testosterone has been postulated to be involved in ALS causation. Materials and methods CSF levels of free testosterone and dihydrotestosterone were measured in 13 ALS patients [7 males, 6 females] and 22 controls [12 males, 10 females]. Results CSF free testosterone levels did not show any significant differences but CSF dihydrotestosterone levels were significantly decreased in all male and female ALS patients. Conclusions DHT is probably integral to survival of motor neurons. In patients predisposed to develop ALS, there is possibly a sort of “testosterone resistance” at level of blood–brain barrier [BBB] existing right from birth and is likely the result of dysfunctional transport protein involved in testosterone transfer across the BBB. In these patients, lesser amount of testosterone is able to breach the BBB and enter the central neural axis. Lesser amount of testosterone is available to 5 α reductase in the anterior pituitary to be converted to DHT and lesser amount of DHT is generated. There is inadequate negative feedback suppression of LH at the level of anterior pituitary by DHT. As a result of higher LH levels, testosterone levels rise in the peripheral testosterone fraction [the fraction outside the BBB] and this explains the various physical attributes of ALS patients like lower Ratio of the index and ring finger lengths (2D:4D ratio), increased incidence of early onset alopecia etc. This deficiency of DHT leads to motor neuron death causing ALS.

. Sparing of neurons of cranial nerves III, IV, and VI and Onuf's nucleus in ALS that lack androgen receptors and are not involved even in advanced MND (Weiner, 1980). A type of MND-X-linked spinobulbar muscular atrophy (Kennedy's disease), results from a trinucleotide repeat expansion in the androgen receptor gene (La Spada, Wilson, Lubahn, Harding, & Fischbeck, 1991). Athletes especially those engaging in contact sports like Soccer/baseball which require a high degree of endurance have an increased susceptibility to ALS. Studies have found that both male and female athletes in aggressive contact sports have higher testosterone levels (Pillay, 2006). Some authors have hypothesized that individuals with high prenatal testosterone exposure self-select into aggressive sports and occupations which require endurance (Wicks, 2012). Fondell, Fitzgerald, Falcone, O'Reilly, & Ascherio (2013) found an increased risk of ALS in men with early-onset androgenic alopecia. Gargiulo-Monachelli (2014) found that postmenopausal female ALS patients had significantly higher serum total testosterone and serum free testosterone concentrations than age-matched postmenopausal controls. Studies have found a link between head trauma and ALS Chen, Richard, Sandler, Umbach, & Kamel (2007). This can be explained by the simple fact that many of these injuries occurred in those engaging in contact/collision sports and as mentioned in point 4, probably individuals with high testosterone self-select into aggressive sports/jobs requiring physical exercise. Vivekananda et al. (2011) found that ALS patients had lower Ratio of the index and ring finger lengths (2D:4D ratio) in comparison with controls. 2D:4D ratio is dependent on prenatal testosterone levels. A study (Militello et al., 2002) of male and female patients of ALS found that serum free testosterone levels were significantly lower in both male and female ALS patients.
However, there also have been observations which did not directly implicate testosterone Bruson et al. (2012) analyzed Androgen Receptor CAG expansions in 336 patients with ALS and 100 controls and found no significant difference between the 2 groups. Jones, Riley, & Antel (1982) treated male ALS patients with high dose testosterone and found that exogenous high dose testosterone therapy caused a predictable decrease in basal LH and FSH levels and expected dampening of LH and FSH response to GnRH stimulation.
In our study, we measured concentrations of testosterone and its principal metabolite dihydrotestosterone in cerebrospinal fluid (CSF) of ALS patients and compared it with those of normal controls.

| ME THODS
CSF levels of free testosterone and dihydrotestosterone were measured in 13 ALS patients [ 7 males and 6 females] and in 22 controls [12 males and 10 females]. All CSF samples were collected in the morning between 9:00 a.m. and 11:00 a.m. as some studies (Goodman, Hotchkiss, Karsch, & Knobil, 1974) have shown that testosterone concentrations vary during the day. Due clearances from Institutional Research and Ethical Committees was obtained.
Written informed consent was taken from patients and controls for CSF and clinical data collection.
Inclusion criteria used for patient selection: a) Patients fulfilling the diagnosis of clinically definite ALS and clinically probable ALS as per the El Escorial Criteria (EEC) (Brooks, Miller, Swash, & Munsat, 2000).
Inclusion criteria used for control selection: a) Controls were enrolled from surgery, gynecology and obstetrics and from orthopedics wards. CSF was obtained from male and female controls undergoing lumbar puncture for spinal anesthesia for surgery.
Exclusion criteria used for both patient and control selection: CSF concentrations of free testosterone and dihydrotestosterone were measured using solid phase enzyme-linked immunosorbent assay (ELISA) based kits sourced from IBL International.

| RE SULTS
Clinical and demographic data and CSF concentrations of free testosterone and dihydrotestosterone in ALS patients and controls are summarized in Tables 1-5 and Figure 1.
The data sets were found to have a normal distribution using

| D ISCUSS I ON
In our study, we have demonstrated significantly decreased CSF dihydrotestosterone levels in both male and female ALS patients.
More importantly, all ALS patients studied had decreased CSF dihydrotestosterone pointing to this being an important contributing factor to ALS pathogenesis.
There have been few studies done on CSF penetration and CSF metabolism of testosterone and dihydrotestosterone and how testosterone and its metabolites exert a negative influence on LH release. These studies have demonstrated the following observations.
Radioactively labeled dihydrotestosterone was given intravenously to six adult male rhesus monkeys (Macaca mulata), and it was found that that almost no radioactively labeled DHT could be found in CSF postinjection thus demonstrating that dihydrotestosterone has minimal penetration across the blood-brain barrier [BBB] (Marynick, Havens, Ebert, & Loriaux, 1976). A study (Abbott, Batty, Dubey, Herbert, & Shiers, 1985) on castrated male monkeys found that even in presence of very high, supraphysiological serum levels of DHT, CSF concentrations of DHT stayed very low. Also, there was no correlation between levels of unbound fraction of DHT in serum and DHT in CSF despite the fact that in CSF, the entire DHT fraction is unbound. The authors hypothesized that some special mechanism or carrier protein limits the influx of DHT in CSF from serum even in presence of very high supraphysiological DHT serum concentrations (Abbott et al., 1985). In another study (Schaison, Renoir, Lagoguey, & Mowszowicz, 1980), it was found that administration of DHT was unable to suppress LH in either normal men or agonadal patients.
DHT was administered through the percutaneous route as a hydrosoluble gel for 3 months to all subjects and plasma DHT levels 8-10 times the normal plasma range were reached and maintained without any reduction in circulating LH levels. This study demonstrated that DHT is unable to enter the CSF compartment despite the presence of sustained high, supraphysiological serum levels. The above findings are in contrast to CSF penetration dynamics of testosterone. Studies (Backstrom, Carstensen, & Sodergard, 1976;Dubey, Herbert, Abbott, & Martensz, 1984)  levels with the entire CSF testosterone being unbound and that there is a predictable relationship between serum total testosterone levels, serum unbound testosterone levels and testosterone levels in the CSF. Also, it is known that testosterone crosses the BBB (blood brain barrier) only in the unbound state. These studies also found that serum testosterone concentrations show a pronounced diurnal variation and that in castrated animals, concentrations similar to the high early morning serum concentrations in the precastrated state could only be achieved with much higher doses of testosterone. The study (Dubey et al., 1984) also found that a sharp increase in serum total testosterone, unbound serum testosterone and CSF testosterone coincided with an abrupt fall in serum LH (Luteinising hormone) levels. Abbott et al. (1985) demonstrated in their study that DHT can suppress serum LH levels and also concurred with a previous study by Sholl, Goy, & Uno (1982) that a majority of DHT within the brain comes from the precursor testosterone. A study (Martini, Celotti, & Melcangi, 1996) using fetal and neonatal rat brain cells found that formation of DHT takes place preferentially in the neurons in the nervous system although type-2 astrocytes and oligodendrocytes also possess some 5 alpha-reductase activity. Other studies (Celotti, Melcangi, Negri-Cesi, & Poletti, 1991) have also echoed similar findings. Pardridge, Moeller, Mietus, & Oldendorf (1980) in their study found that DHT is sequestered to a greater degree in mammalian brains than is testosterone. Moreover, it is a known fact that DHT is a much more potent hormone than testosterone because its binding affinity to the androgen receptor is two times that of testosterone and it has a dissociation rate about a fifth of that of testosterone (Marchetti & Barth, 2013). Sholl, Robinson, & Goy (1975) from their study concluded that DHT is the primary androgen for activation of neural mediated effects, at least in the guinea pig. In a study done by Zoppi et al. (1988) on rats, it was found that when testosterone was given along with an inhibitor of 5 alpha-reductase thereby leading to lesser conversion of testosterone to DHT, LH levels were reduced to a lesser extent than they were when testosterone was given alone.
Also, it has been found that patients with 5 alpha-reductase type 2 deficiency have high circulating LH levels despite having normal or elevated serum levels of testosterone. However, these patients have reduced serum DHT levels (Martini, Celotti, & Serio, 1979). Various other analyses (Zanisi, Motta, & Martini, 1973) have also shown that DHT and its metabolite, 3-alpha diol are much more effective that testosterone in suppressing LH secretion. A study evaluating effects of testosterone,DHT and its metabolites in cultured pituitary cells (Denef, 1983) provided further evidence for a specific physiological role of 5-DHT and 3 -alpha diol. This study concluded that at least at the gonadotroph cell level, DHT and possibly 3-alpha diol are the active androgens which depress LH release. Another study (Kennedy, Rawlings, & Cook, 1985) done on bull calves also found that LH serum concentrations were suppressed by administration of sialistic implants releasing DHT and also by 3-alpha diol implants. Studies (Martini, 1982) have found that the anterior pituitary very efficiently metabolises testosterone to DHT and 3α-diol. Yields of DHT generated are second only to yields seen in the prostate gland. Also, in the anterior pituitary, around 25% of testosterone is converted into 3-alpha diol while in the prostrate, only 7.6% of testosterone is converted into 3-alpha diol. It has also been observed that the anterior pituitary of adult mammals does not have significant aromatizing ability while the hypothalamus has potent aromatizing abilities.
Studies (Chimento, Sirianni, Casaburi, & Pezzi, 2014) suggest that estradiol derived from intracerebral testosterone is the main hormone that provides negative feedback at the hypothalamic level while at the level of the anterior pituitary, both DHT and estradiol, both derived from intracerebral testosterone, are required for negative LH feedback. Numerous animal studies (Abdelgadir et al., 1994;Scott, Kuehl, Ferreira, & Jackson, 1997) also support that estradiol derived from intracerebral testosterone is the main hormone that provides negative feedback at the hypothalamic level while at the anterior pituitary level, DHT also plays a very important role in suppressing LH. Interventions like reducing DHT levels with either a reductase inhibitor or by using antibodies to estradiol or through administration of aromatase inhibitors-all these maneuvers compromise the ability of T to suppress LH.
Interestingly, the anterior pituitary and the hypothalamic/preoptic area possess 5 α reductase type 2 isoenzyme whereas the glial cells and the neurons predominantly express the 5 α reductase type 1 isoenzyme. It was also found that in the rat pituitary, 5 α reductase type 2 isoenzyme is located mainly in the gonadotropes.
Various authors have concluded that type 1 5 α reductase is constitutively expressed in the mammalian brain and appears to have a neuroprotective role and mutations in type 1 5 α reductase lead to death of the organism in gestation itself ( highest levels being seen in the morning and lowest in the evening (Bremner, Vitiello, & Prinz, 1983;Plymate, Tenover, & Bremner, 1989;Tenover, Matsumoto, Clifton, & Bremner, 1988). A study (Bremner et al., 1983) documented that elderly males have reduced testosterone levels as compared to younger males but more importantly, there was a greater, more pronounced difference between the circadian excursion of total serum testosterone levels between the two groups. Plymate et al. (1989) found that elderly males have a reduced significant circadian rhythm in free non-SHBG bound testosterone as compared to younger males [60% versus 100%]. This study further found that even in 60% of older males showing circadian variation in free non-SHBG bound testosterone, the circadian excursion was only 26% of what was seen in the younger males. Tenover et al. (1988) found that the circadian pattern of pulsatile LH secretion was blunted in healthy, elderly males as compared to young males.
Studies (Gapstur et al., 2002;Leifke, Gorenoi, Wichers, Von Zur, & Brabant, 2000) have consistently shown that both total testosterone and free testosterone concentrations decrease with ageing while SHBG levels increase with ageing. Studies (Laaksonen et al., 2004) have shown that obesity, insulin resistance, metabolic syndrome and dyslipidemia have a strong association with low serum levels of total testosterone, free testosterone and sex hormone binding globulin (SHBG). In addition, it was found (Laaksonen et al., 2004)  This erroneous assumption probably stemmed from the higher detection limits of the hormonal assays used in the previous studies (Mitamura et al., 1999). However, studies using ultra-sensitive assay itself cannot penetrate BBB. Dubey et al. (1984) had found that a sharp increase in serum total testosterone, unbound serum testosterone and CSF testosterone had coincided with an abrupt fall in serum LH (Luteinising hormone) levels. We postulate that when high testosterone concentrations are reached and they breach a certain "critical" threshold, the binding capacity of SHBG (Sex hormone binding globulin) is exceeded and only then does the levels of free, unbound serum testosterone rise sufficiently to cause the rise in CSF testosterone [ in CSF, all of testosterone is free and unbound]. This CSF testosterone is converted to dihydrotestosterone and the resultant increase in dihydrotestosterone suppresses LH levels by exerting a negative feedback on LH release. We postulate that in patients who are predisposed to develop ALS, there is a sort of "testosterone resistance" at the level of BBB. This "testosterone resistance" exists right from birth. This "testosterone resistance" is likely to be the result of a faulty, mutated transport protein involved in testosterone transfer across the BBB. In these patients, lesser amount of testosterone is able to breach the BBB and enter the central neural axis. As a result, lesser amount of testosterone is available to 5 α reductase type 2 isoenzyme in the anterior pituitary to be converted to DHT and lesser amount of DHT   (Bruson et al., 2012) found normal androgen receptor function in ALS. More studies evaluating these parameters and also concurrent serum testosterone, serum dihydrotestosterone and LH levels would shed more light on ALS pathogenesis.
Conclusions-Our study implicates dihydrotestosterone deficiency in the neural axis as an important component of ALS pathogenesis. More studies evaluating these parameters and also concurrent evaluation of serum testosterone, serum dihydrotestosterone and LH levels would shed more light on ALS pathogenesis.
Long-term serial studies on asymptomatic familial ALS gene carriers would be invaluable in delineating the pathogenesis of this dreaded disease.

ACK N OWLED G EM ENTS
Prof. Narinder Kumar, Professor and Manish Goyal , Research Fellow, Department of Statistics, Panjab University ,Chandigarh for performing the statistical analysis in the study.

CO N FLI C T S O F I NTE R E S T
None. Controls for the study. There is no conflict of interest for any of the authors. The authors take full responsibility for the data, the analyses and interpretation, and the conduct of the research; full access to all of the data; and the right to publish any and all data.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.