The physiological role of testosterone in erectile function

Clinical experience with testosterone treatment in Sildenafil non-responders

Abdulmaged M. Traish, PhD, Professor of Biochemistry and Urology, Boston University School of Medicine

I would like to discuss the potential role of testosterone in sexual function, specifically in sexual arousal function in the male. More specifically I will discuss the physiological effect of testosterone on penile tissue structure and cellular signaling. Although there is ample data and studies on role of testosterone in sexual function, there are still a number of issues that need to be better understood. For instance, what does androgens do in the male penis, and how erectile function is dependant, or not dependant, on androgens.

There are several factors that modulate erectile function. There are vascular issues, there are neurogenic components, but what I want to focus on today is the hormonal component. As we begin to appreciate, the hormonal component is not very simple, because you have a whole slew of hormones that might affect penile tissue. Also we have to keep in mind that penile injury and penile diseases play an important role in erectile dysfunction. We also cannot forget that we are in the Aging Male meeting here, and many issues pertain to medication and we have to keep that in view.

The point I want to make, and I hope that I can make it clearly today, is that the endocrine milieu of the individual plays an important role in erectile function. The interplay between structural component of sexual function organs, as well as the neural component modulates the vascular component. So we have an integration of systems that come into play, which eventually modulate male erectile function. Any alteration in the structural , vascular or the neural components will modify the physiological function. Thus, hormonal alterations (loss of androgen) may ultimately lead to erectile dysfunction.

Recognise that there is a lot of controversy whether testosterone is important or not important in erectile function. If you look, there are quite few studies in the literature. You can go back as far as 1963 or even earlier, up to today, and these studies suggest that medical, or surgical castration, can result in loss of libido and/or erectile function. To this day, this issue is being debated and I hope I will shed some light today on what are some of the main issues.

If you look at either Massachusetts Male Aging Studies, or studies by Rhoden, they both suggest that androgen levels drop in aging and androgen concentrations are low in the serum, but there is no direct relationship between androgens and erectile function. I think this is an important point to keep in mind because we think of hormones and penile erectile function as immediate response rather than in the sense of maintaining organ cellular structure and signaling function.

If you think by taking testosterone tablets or gel, and within two hours, or 15 minutes, or 45 minutes, you are going to get an erection, that’s not what we’re talking about. We are talking about maintaining vascular structural integrity of the penile tissue namely, the corpora cavernosa.

So what I would like to do today is show you some of the data from our animal model studies. In this animal model studies we did two things; we did either surgical castration or medical castration. We focused on looking at some cellular signaling and the fibroelastic properties of the tissue, and how androgen modulates these two particular properties, and eventually anything that modifies the corpus cavernosum compliance. will lead to veno-occlusive dysfunction. Once you have veno-occlusive dysfunction, you have venogenic impotence and therefore you will have erectile dysfunction.

Upon sexual stimulation several things happen. First you have to have vasodilation of the helicine arterioles and resistance arteries and an increase in blood supply due to arterial dilation. This process is followed by trabecular smooth muscle relaxation to allow the tissue to expand, and permits venous compression to impede blood outflow. So androgen may affect changes in blood flow either by affecting the function of the smooth muscle and by maintaining the fibroelastic properties of the tissue, or androgens might affect the signaling pathways modulating trabecular smooth muscle relaxation. Thus, any of these components could be altered by androgen insufficiency or deficiency. .

Outflow is impeded upon sexual stimulation and we have visible dilation of the penile artery and the helicine arterioles. There is increase in blood inflow, there is decrease in blood outflow because of the compression of the subtunical venules due to the expansion of the corpus cavernosum against the tunica albuginea. Therefore, expansion of the tabecular smooth muscle, or relaxation of the trabecular smooth muscle and allowing the lacunas vessels to fill with blood, the penis traps blood under pressure, and acts as a capacitor.

In a microscopic fashion there are several components warrant to be highlighted. You have the endothelium lining the lacunar spaces, you have the trabecular smooth muscle and of course the penis has to sustain the pressure of the penetration into the vagina, therefore it has to have rigidity, and that rigidity is maintained by the fibroelastic properties of the tissue. So all penile structure components must be functional. The metabolic integrity must be maintained in order for the tissue to function properly. Androgen may affect any of these metabolic and functional properties of penile tissues structural elements. . It may affect endothelial function, it may affect smooth muscle function, it may alter the balance between trabecular smooth muscle and connective tissue.

This is a slide I have taken from Shipley & Dorfman (1956) book on Androgens, and it tells us a very simple thing; it tells us that androgen are very important for development and maintenance of genital tissue. As we can see here, as we go from nine, to ten years old, to 18 years old, we can see penile tissue grows in size. The testicles also grow, and therefore androgens are important for development, as well as maintenance, of genital tissue function integrity.

What we’ve done is a very simple study. We took two groups of animals, a small group, small means per weight, these are roughly five to seven months old and another group which we call large, between 10 and 14 months old, and we did a very simple experiment. We took these animals and we tried to measure intra-cavernosal pressure subsequent to pelvic nerve stimulation. You can see in the young animal, although we stimulate the pelvic nerve, we do not trap blood very easily in the corpus cavernosa, either at 2Hz or 4Hz. However, in the older animal or mature animal in this case, we can get a nice increase in the intra-cavernosal pressure and when we just quantify that, you can see that two and 4Hz, does very little entrapment of blood into the very young animals, but in the mature animals there is a wonderful increase in the intra-cavernosal pressure. At high frequency we can see an increase in the young animals as well. .

One of the questions we thought at the time was that, if this is to do with androgen maintenance of the sexual organ, in this case, the corpus cavernosa in the penis then we should see it. Mason trichrome staining of the tissue reveals interesting information. You can see in the immature or young animals, although all the cellular elements are there, (red stain means smooth muscle and blue stain means connective tissue), but if you can see here in the immature animals, the lacunar spaces are smaller and that probably means the endothelium may not be functional with regard to signaling . But if you look at the tissues sections from mature animal you can see that the lacunar spaces are large and also, is the tunica albuginea fibrous arrangement is different (which is not very clear here from this slide). This suggests that the entrapment of blood in the penis and the veno-occlusive mechanism requires functional integrity, and this functional integrity is maintained by androgens.

This is the same tissue section, only stained with elastin specific staining and we can see quite a bit of elastin staining in the mature animal vis-à-vis the young (immature) animals.. Unfortunately I do not have quantification of this staining here.

In the next few minutes I want to focus on how androgen affects blood flow. How androgen affects signaling, vis-à-vis alpha adrenergic receptor? . How androgen affects nitric oxide synthase in the endothelium, as well as the neural nitric oxide? How androgen affects PDE type 5?. And the most important point, how androgen affects the trabecular smooth muscle content, and the connective tissue matrices and how this balance is clearly critical for erectile function? And androgen deficiency may alter this balance dramatically, androgen deficiency or androgen insufficiency, might be a very important point of targeting. Of course, there are studies, not from our laboratory, but from other groups, which shows that androgen alter nerve fibre density as well as endothelium function.

Here are the two models. This is the surgical castration model, you can see that the androgen levels dropped significantly. Here is the medical castration model by treating the animal with the LH/RH agonist, and you can see the decrease in plasma testosterone is significant, but not completely to the surgical model. Although there is only a small, but considerable drop in testosterone levels, not as much as noted in the surgical model, you will see from the data that, both as long as you have a decrease in testosterone levels you will see an immediate effect on the veno-occlusive mechanism as determined by intracavernosal pressure in response to pelvic nerve stimulation. .

If you look at the control animals, this is the intracavernosal pressure, there is a sharp increase in the intra-cavernosal pressure. If you surgically castrate the animal you will obliterate this process, this means the animal has a venous leak or veno-occlusive dysfunction. If you treat these castrated animals with testosterone you can restore this function. However, if you treat them with estradiol only, you do not restore this function suggesting that this is an androgen-dependent response.

If you normalize the data, there is no effect on the systolic or systemic arterial pressure, but you can see here the control, surgically castrated, testosterone-treated after surgical castration or oestrogen treated you can see that there are considerable changes in the intracavernosal pressure in response of electric stimulation of the pelvic nerve as compared to the flaccid state. So, clearly androgens modulate erectile function in the sense that they maintain the veno-occlusive mechanism.

This data is from the medically castrated animal, but this slide is a surgically castrated model. You can see this is the control, this is the surgically castrated animal, and you can see that there is a marked degree in the intracavernosal pressure at all frequencies. Two weeks after Lupron treatment, four weeks or eight weeks, you can see a marked progression in the decrease in the intracavernosal pressure and the decrease is significant Whether you have medical or surgical castration, any means by which you diminish the testosterone signal, you have physiological impact on entrapment of blood vis-à-vis the veno-occlusive function.

One key point I want to make here and I hope I don’t make it ambiguous. This is the control animal, if you surgically castrate, or medically castrate, you cannot restore erectile function by giving PDE 5 inhibitor alone. Once you alter the tissue fibroelastic properties, you will not be able to regain the function by PDE 5 alone because this is not just the NO pathway, but rather the tissue architecture, the cellular composition with additional change and you cannot restore or correct the veno-occlusive dysfunction in this case, by giving PDE 5 inhibitors alone.

When we look at the signaling, as you can see here, the control animal when you castrate, there is considerable decrease in α1-adrenergic receptor, if you treat with testosterone, you restore it. This is telling us that the smooth muscle function has changed, probably from a contractile into synthetic, or that the smooth muscle change in itself may reflect that androgen component.

This is normalised based on DNA so it’s not normalised by protein so that it takes into consideration the apoptosis in the smooth muscle cell.

If we look at the neural nitric oxide synthase in this animal model, we do not see marked changes by castration. So in this case, somehow, neural nitric oxide synthase as measured by expression with western blots. You don’t see much change between the castrated, or castrated, testosterone-treated. Why? We don’t know. In the rat this seems to be the case. If we look at the activity of nitric oxide synthase you can see that we don’t see the changes by castration. This is control , castrated, castrated and estradiol -treated, or castrated and testosterone treated, we don’t see really marked change in the nitric oxide synthase in these animals. If we normalize the activity data to DNA, we see the same thing, we don’t see any significant change, so this is not due just to quantitation, but rather we do not see changes in the neural nitric oxide synthase in this model. However, we see changes in the endothelial nitric oxide synthase. We can only do that by immunohistochemistry. We were not successful in doing this with western blot analyses.

This is now from the Lupron treated model. It’s the same activity. We measure the activity in control and medically, or surgically castrated animals. In the medically castrated animals, we have two weeks after Lupron, four weeks, eight weeks, you can see there is no change in the total nitric oxide synthase activity, so it seems this is not the target, at least in our view here in this model.

If we look at the PDE 5 it is interesting, there is a very small drop but not very significant, but after treatment of the castrated animal with androgen, we can see a significant change in the PDE type 5, and this suggests that PDE type 5 is an important regulator in that pathway. And maybe by giving testosterone and u normalizing the function of tissues, increases in PDE type 5 activity may be important to protect against potential priapism in this case, and therefore androgen may up-regulate PDE 5. We have not really understood this yet but we are still doing some work in cell culture to look at this in much more detail.

Here is probably the key point of focus. If you look at the Masson Trichrome staining of penile corpora cavernosa from the intact animal, you can see the red represents smooth muscle and blue represents the connective tissue matrix, and upon castration whether surgically, or medically, you can see that there is a marked decrease in the content of smooth muscle and there is an increase in the connective tissue matrix. If you treat the castrated animal with androgen, you restore it almost to normal. If you treat it however, with oestrogen, you can see there is quite a bit of connective tissue, the smooth muscle content was not restored to normal..

If we quantify this by computer assisted histomorphometerey, you can see that in the control, no change in smooth muscle. Upon castration you can see there is a 20 – 23% decrease in the smooth muscle content. You cannot reverse or restore such decrease by oestrogen treatment, however you can reverse this by testosterone treatment, suggesting that loss of the smooth muscle will result in loss of expandability, and loss of compliance, and therefore veno-occlusive dysfunction, as shown by the physiological data on intracavernosal pressure measurements.

This is just another set of data from the Lupron model, and as you can see the impact when you compare the surgically orchiectomised, and the Lupron, you can see there’s quite a bit of the connective tissue matrix, in the androgen-deficient animals compared to the control animals. This observation clearly suggests both surgical, or medical, castration result in alteration in smooth muscle connective tissue balance, and this balance is important. If it’s altered you will end up with veno-occlusive dysfunction, and therefore, erectile dysfunction.

This is just to come back to the point that, although testosterone levels here are higher than here, medical versus surgical, it’s still when we look at the intracavernosal pressure, or we look at the tissue cellularity, we see similar changes in both models.

These two pieces of data are from Drs. Rogers and Lue (San Francisco). What you can see here on the slide, this is an intact rat tissue and you can see this is tissues from castrated rats. Upon castration you can see there is quite a bit of atrophy of the smooth muscle cell, and there is deposition of connective tissue proteins, and the smooth muscle cells are far apart. If you treat the rats either with testosterone or VEGF in this experimental model, you go back from altered smooth muscle connective distribution, into a more normal distribution. This is just a high magnification (slide).

What’s interesting in his model is that the authors also show that if you take the intact animal and look at the myelinated and non-myelinated nerve fibres, here, after castration the diameter as well as the thickness of the myelin sheath is reduced by castration and in this case if he animals were treated with VEGF, instead of androgen, the myelin sheet as as well as the thickness of the dorsal nerve diameters are normalized. In a personal discussion with Dr. Lue this past October he said they can reproduce this either by giving testosterone, or by giving VEGF, suggesting that testosterone might be acting via a paracrine mechanism involving VEGF. I am coming to summarize.

So, simply the erectile function mechanism is a very integrated process that requires input from the neural component, the hormonal component, structural component, vascular component and when all is normal you can go from the flaccid state to the erect state upon sexual stimulation, because you have an operating veno-occlusive mechanism.. However, if androgen deficiency and androgen insufficiency settles in, you will alter signaling, as well as tissue cellularity and tissues architecture. You decrease total trabecular smooth muscle, you increase the connective tissue, you alter the fibroelastic properties of the tissues, and therefore you pose what we call venogenic or venous leak, and you reduce erectile dysfunction.

In summary, in the animal model, and I have to be very specific, I say in the animal model, androgen insufficiency produces decrease in alpha-adrenergic receptor expression. We did not note any marked changes in the neural nitric oxide synthase in this animal model, (i.e. this rabbit). However, we see some changes in the endothelium nitric oxide synthase. We noted that androgen treatment of castrated animal increases PDE activity. This is an observation that we have yet to clearly understand its significance. But what is really most important is that there is a great deal of reduction in the trabecular smooth muscle content, and function. And reduction of smooth muscle content, coupled with increased connective tissue matrix, the deposition of this matrix alters the fibroelastic properties of penile tissue compliance and this together, results in a veno-occlusive dysfunction. Therefore this is measured physiologically by a decrease in intracavernosal pressure.

The final message is that structural and functional alteration in the corpora cavernosa produced by androgen deficiency results in corporal veno-occlusive dysfunction. The changes produced by androgen insufficiency cannot be reversed with treatment using PDE 5 inhibitors only. If you have veno-occlusive dysfunction which is based on changes in tissue fibroelastic properties, you are not going to correct that by giving PDE 5.

And ultimately we have to put all this in context from the literature. What we suggest is that under normal androgen levels you have synthesis of growth factors, connective tissue growth factors, metallporteases, vascular endothelium growth factors which are regulated by androgens, which in turn regulate connective tissue synthesis and deposition in the penis. You have connective tissue growth factor which is also regulated by androgen. Drs. Lue and Rogers data suggests that androgen up-regulate vascular endothelial growth factor (VEGF). The effects of androgens via paracrine mechanisms via VEGF or directly via receptor mediated mechanism is important to maintaining tissue fibroelastic properties, smooth muscle content and function and maintenance of physiological function and ultimately erectile function. Any hormonal imbalance (androgen insufficiency or androgen deficiency), will modify these physiological and biochemical pathways. Why, because it will alter the nerve structure and function, vascular smooth muscle as well as the endothelium and will result in veno-occlusive dysfunction and will result in erectile dysfunction.

I want to conclude by acknowledging and thanking my colleagues Dr. Noel Kim who is a physiologist and has been with us for almost ten years, Dr. Huang, our technician, Jennifer Gottlieb and I would also like to thank Dr. Seong Choi, Kweonsik Min, Soo Woong Kim and Seong-Joo Jeong and also Dr. Kwangsung Park who is in the audience here today who also participated in this study and also my colleagues, Drs. Irwin Goldstein and Ricardo Munarriz. Thank you very much.