Sellami H, Gdoura R, Mabrouk I, Frikha-Gargouri O, Keskes L, Mallek Z, Aouni M, Hammami A. A proposed mouse model to study male infertility provoked by genital serovar E, Chlamydia trachomatis. J Androl. 2011;32: 86–94.
The study of Chlamydia trachomatis (CT) infection in the male genital tract has been the focus of attention of many researchers during last 2 decades. However, it still is a controversial issue to be clarified. Many aspects of this topic remain inconclusive. Is CT a definitive uropathogen affecting the upper genitourinary tract? Does this infection affect semen quality and/or male fertility? Is there a valid animal model to study male urogenital infection by CT? Which strain of CT is adequate to use in an animal model to mirror what happens in a human male urogenital infection?
We would like first to report our own experience. We have been studying chlamydial infection in the male genital tract in the human disease chronic prostatitis/chronic pelvic pain syndrome (CP/CPPS) and in developing experimental models of this infection in rodents. By studying the role of CT in CP/CPPS, we have found that CT infection has an occurrence of approximately 15% in the CP/CPPS patient population and that this infection does not affect semen quality; thus, it would not impair male fertility (Motrich et al, 2005, 2006a). To go further, we developed 2 animal models of CT urogenital infection in rats and mice. We infect rodents with Chlamydia muridarum in the meatus urethra, the natural route of infection, and animals gradually develop an infection primarily present in many regions of the genital tract (urethra, bladder, seminal vesicles, prostate, epididymis, and testes); however, when it then becomes a chronic infection, bacteria show a special tropism for the prostate tissue because this gland is the main reservoir of the infection. This infection, either in the acute or chronic phase, does not affect semen quality (sperm density, viability, and motility) and fertility potential (Potency, Fecundity, and Fertility Index) of infected animals (Mackern-Oberti et al, in preparation).
Some aspects of the work by Sellami et al sparked our attention. First, we have concerns regarding the CT strain used to infect animals in the meatus urethra. It is well known and widely accepted that humans are the exclusive natural host for CT. Rodents are naturally infected with other species of Chlamydia (eg, C muridarum) (Miyairi et al, 2010). In fact, the work done by Pal et al (2004) is cited many times, which is one of the first studies to achieve a mouse model of male urogenital infection. In this model, Pal et al (2004) used Chlamydia trachomatis mouse pneumonitis biovar, also known as C muridarum. Mouse models of chlamydial genital tract infection use common inbred mice strains and C muridarum. It is the most commonly used model because of the reproducible disease course, cost, and ability to manipulate immune parameters (Miyairi et al, 2010). Although mice can be infected urogenitally with human CT serovars, a higher inoculum is required with CT than with C muridarum, the peak bacterial load is 1 to 2 logs lower, the infection is eradicated relatively quickly, and pathology is mostly limited to the lower genital tract (Miyairi et al, 2010). Because of these facts, the validity of a mouse model that uses a CT strain that naturally does not infect mice is compromised.
Secondly, chlamydial infection was only analyzed in the testes and epididymis. In our opinion, it would be necessary to assess chlamydia's presence in all of the main parts of the male urogenital tract (eg, urethra, seminal vesicles, prostate, epididymis, and testes) to propose an animal model to study male infertility provoked by CT.
Third, in the article, it was pointed out that semen parameters were evaluated. However, after we carefully read the article, we noted that every assay performed to test sperm quality was done using epididymal sperm. As is well known, sperm are produced in the testes and stored in the epididymis until the time of ejaculation when they join sexual accessory gland secretions and form semen. To study semen, we have used/developed 2 methods for obtaining samples from rodents (Motrich et al, 2006b, 2009). In our experience, it is not equivalent to study one sample or the other, and the rationale for this is that semen is not only composed of sperm cells. Infections or inflammatory mediators, such as reactive oxygen species present in gland secretions can affect sperm cells, even though they are perfectly normal in the testis and epididymis. When they come into contact with these mediators/infectious agents during and after ejaculation, sperm damage can start to occur (Motrich et al, 2006b). Taking these data into account, it would be necessary to assess semen sperm quality to elucidate if CT has deleterious effects on sperm quality from infected animals. On the other hand, Sellami et al only found significant impairment of epididymal sperm when Chlamydia was no longer detectable (day 30 after infection), an interesting finding that was not discussed.
Fourth and most important, we have many doubts regarding the Reproductive Performance Study. In this section of the work, the fertility index was calculated as the total number of infant mice/total number of pregnant females. With this formula, the variation in the number of oocytes produced by ovaries from every female is not taken into account. In other words, the fertility index calculated in this way is not normalized. In our opinion, a more proper and accurate formula for the fertility index is calculated as the ratio of the number of implantation sites/number of corpora lutea; thus, variations in ovulation capacity by each female under study are taken into account. Then the only parameter that influences the fertility index is the sperm capacity to fertilize an oocyte (Motrich et al, 2009). It would be interesting and helpful to also calculate the preimplantation loss ([number of corpora lutea 2 number of implantation sites]/[number of corpora lutea] × 100) and the postimplantation loss ([number of implantation sites 2 number of live fetuses]/[number of implantation sites] × 100). On the other hand, and a very important point, comparisons were only made between 2 animals from the infected group with 2 control animals at each time point of the experimental protocol. We have concerns of how significant these results may be taking into account the small sample size (number of animals in each group). In our opinion, a sample size of at least 3 animals per experimental group should have been analyzed.
CT infection and its relationship to sperm quality and male infertility is a field that fortunately is the focus of attention of researchers. Proper assessment of the relevant topics and valid experimental models are promptly required to shed light on the many controversial issues of this infection.