People who receive aggressive treatment for certain types of cancers may have fertility problems, and children and teenagers “are often of special concern,” according to the American Cancer Society.
Treatments like chemotherapy and radiation therapy can damage sperm-forming cells and result in impaired spermatogenesis, the origin and development of sperm cells within the male reproductive organs.
Israeli scientists led by a research group at the Ben-Gurion University of the Negev (BGU) in Beersheba have developed an innovative platform to create sperm in a laboratory through a microfluidic system, which contains hundreds of microchannels for fluids to pass through.
The sperm was grown on a special silicon chip developed in collaboration with researchers at the Technion – Israel Institute of Technology. A 3D system was built and integrated to allow the addition of testicular tissue cells. The chip enabled the researchers to grow cells from the testis in the microchip and add fresh cell culture media designed to support cellular growth.
Their research was published recently in the peer-reviewed journal Biofabrication under the title, “Testis on a chip — a microfluidic three-dimensional culture system for the development of spermatogenesis in-vitro.”
Prof. Mahmoud Huleihel from the Shraga Segal Department of Microbiology, Immunology, and Genetics, in the Faculty of Health Sciences at Ben-Gurion University tells NoCamels there was a need to find a method to produce sperm cells in the laboratory, especially for younger males who have not developed sperm yet.
According to Huleihel, when his team began studying some 15 years ago, they knew that when cancer affects sperm generation in adults “medical professionals can cryopreserve sperm from those patients and use them in the future before chemotherapy treatment,” he explains, “The problem is with prepubescent boys who do not generate sperm at their age and the question was how can we cryopreserve the sperm, or how we can make preservation of their fertility if it’s possible.”
Like adult males, these younger males have stem cells, or what he refers to as basic germ cells, which can proliferate, propagate, and differentiate later to produce sperm. “These basic cells are present in boys. So the question was, can we use these cells with any kind of technology to produce sperm in the future?”
Young mice that didn’t produce sperm cells yet were a model the researchers used to imitate conditions for the growth of sperm cells in the testicle. Under laboratory conditions, it was possible to develop a procedure to culture testicular cells in an environment similar to a natural environment.
Prof. Huleihel admits that this is something that has been done before by other scientists around the world. The mice were given specific stems cells which in turn produced fertile sperm. In the case of the boys, the group wanted to find a method to produce sperm that bypasses what Huleihel refers to as “limitations” such as the potential return of cancer cells to the patient’s body.
“The problem with the boys is that we were afraid that some of the tissues that we were going to use in the future still have residual cancer cells. So if we inject these cells to the boys, we might restore the cancer,” he says. Today, there aren’t any methods to isolate the stem cells from the cancer cells. So it’s an admittedly big limitation as to why we cannot use some of the technology already successful in the animal model.”
Rather than trying to isolate stem cells from cancer cells, the BGU researchers decided they would grow their own stem cells using their novel testis-on-a-chip (ToC) platform with a complete 3D system that allowed the addition of growth factors, and cells from the testicles or any other cells from the body.
“We thought about some things that we could add in the culture dish to grow with the stem cells,” Huleihel says, “So that we can grow them and use them in differentiation to stem. Then we could isolate these sperm and inject them into a female with limitation of possible cancer restoration.”
The innovative system was successfully tested on sexually young mice that have germ cells for spermatogenesis. The development in the culture was examined after five to seven weeks. Sperm-like structures were found that contain cells in advanced stages (called round spermatids) in the process of sperm formation. The sperm cells reached 95 percent viability, according to the study. The researchers found that the round spermatids eventually fertilized normally and produce offspring.
The research group is now preparing for the next phase of applying the experiment to cells from humans, Prof. Huleihel tells NoCamels. He gives NoCamels a rough estimate of this testing on humans starting in the next two or three years.
“It is very important that this system be optimized using the animal model. And the next step is to use humans to see if we reach the same stage. If the success in the human is similar to what happened in the mouse system — and if they will also become fertile. This is the future stage,” he explains.
He adds, “If we succeed in creating round spermatids, or even complete stem [cells], then we have the patient’s tissue in the future, right? We split this tissue into two things — one that we use for research and the other we keep for the future for these boys.”
The ultimate goal is to use the platform to produce round spermatids that “can fertilize all side, then we can provide it to any hospital,” Huleihel says. “This the aim of science. It’s something that will benefit all of humanity.”
The research group included Prof. Emeritus Eitan Lunenfeld, from the Faculty of Health Sciences at Ben-Gurion University of the Negev and Soroka Medical Center, currently a senior faculty member at Ariel University and Prof. Gilad Yossifon, from the Faculty of Mechanical Engineering at the Technion (currently a faculty member from the School of Mechanical Engineering at Tel Aviv University). The research was led by PhD students Ali AbuMadighem, from the Shraga Segal Department of Microbiology, Immunology and Genetics, Ben-Gurion University of the Negev and Sholom Shuchat from the Department of Mechanical Engineering, Technion – Israel Institute of Technology.
This study was supported by the Israel Science Foundation (ISF) and in collaboration with the Chinese Foundation for Natural Sciences (NSFC) (ISF-NSFC), the Reproduction Hub at the Faculty of Health Sciences, Ben-Gurion University of the Negev, and Council for Higher Education Scholarships for outstanding PhD students from the Ultra-Orthodox and Arab populations.
In Vitro Spermatogenesis
Spermatogenesis is a complex process, whereby spermatogonial stem cells (SSC) divide and differentiate to become haploid round spermatids that eventually undergo spermiogenesis to mature into sperm [79,80]. For many years, scientists have attempted to replicate this process in vitro, with various measures of success. See Huleihel and Lunenfeld , Komeya et al. , and Richer et al.  for comprehensive reviews of the field of in vitro spermatogenesis.
Reproducing the structure and function of the testes in vitro is important for the study of the spermatogenic process and for understanding the cell–cell interactions in the testis and the mechanisms of action of various biological factors that affect testicular function and male fertility. It is also of great clinical importance for pediatric cancer patients facing aggressive gonadotoxic chemotherapy/radiotherapy treatments that can render them sterile. Presently, there are no options available to them to preserve their fertility. Unlike adults, they cannot cryopreserve their sperm as their testes do not yet produce sperm. Current guidelines advise cryopreserving testicular biopsies that contain sperm cell progenitors [84–86]. Often however, the tissue cannot be reintroduced, due to the risk of the presence of residual cancer cells that can potentially reintroduce the cancer [84,85,87].
The ability to induce spermatogenesis in vitro and develop fertile sperm would allow them to father their own biological children via in vitro fertilization (IVF), using intracytoplasmic sperm injection (ICSI). In vitro spermatogenesis could also be used in the treatment of azoospermic patients with maturation arrest or even Sertoli cell only syndrome. Their testicular biopsies show no mature sperm, but often they contain sperm progenitor cells at various stages of differentiation, or in some cases only SSCs, that for various reasons are not progressing through spermatogenesis to mature sperm [85,88,89].
There are generally two key approaches to in vitro spermatogenesis. The first method is organ culture, in which a section of testicular tissue is cultured in vitro. This method was shown to induce and maintain spermatogenesis in testicular tissue from mice for a prolonged period of several months. Sato et al.  was the first to succeed in generating fertile sperm that was then used to produce live offspring. However, this method is restricted by the size and shape of the original tissue section, thereby limiting the ability to incorporate it into an advanced organ-on-a-chip model. This is especially true for humans where only a limited amount of testicular tissue from the patient is available. The proportions and ratios of the various cell types, the geometry, and, to some extent, the microenvironment are also constrained by the composition of the tissue .
The second method utilizes disassociated testicular cells. These cells can be sorted into specific cellular subtypes and cultured, or they can be combined in various proportions and used to form organoids. This method offers greater flexibility; however, although there has been success in replicating some aspects of the in vivo environment, no model has yet
generated organoids de novo that completely replicate both the structure and function of the testis and consistently generate mature and fertile sperm .
Generally, these methods use one of a number of gels, such as Methylcellulose , collagen , agar , Matrigel , alginate , decellularized testis tissue [96,97], or, in some cases, no gel at all . The organoids are cultured in various conditions, such as a well plate, gel droplets , in an air–liquid interface on a block of agar  (see Figure 7C), a bioprinted scaffold , or, in one study, preformed organoids were cultured in a microfluidic chip  (see Figure 7D). Some of these studies have achieved elongated spermatids [91,93,95,100], none yet however, have achieved mature fertile sperm using these methods. Although in one study, using our methylcellulose system, we succeeded in the development of mature sperm, the yield was extremely low and inconsistent, and the fertility of the cells was not proven .
Very few of these studies have incorporated microfluidic technologies. This limits the ability to control the microenvironment of the cells, supply nutrition, and remove waste effectively. See Sharma et al.  for a comprehensive review of the benefits of utilizing microfluidic technology in replicating in vitro spermatogenesis. This review focuses on incorporating perfusion, active fluid flow, and vasculature into in vitro spermatogenesis models and the significant capabilities it provides beyond simpler microfluidic setups. Additionally, strategies used to recreate tubular structures in vitro can be applied to testicu- lar models, for more spatially- and structurally-accurate replications of the seminiferous tubule structure.
Current Microfluidic Technologies Used to Model Spermatogenesis In Vitro
Many techniques have been used to replicate spermatogenesis and the testicular structure in vitro (see the above-mentioned reviews, Huleihel and Lunenfeld , Komeya et al. , and Richer et al. ). Described below are a number of the currently-used techniques that are relevant to the implementation of perfusion-based testis-on-chip (ToC) models.
Using an organ culture method, Komeya et al. and Yamanak et al. cultured testicular tissue fragments from pre-pubertal mice in a microfluidic channel under a membrane  (See Figure 7A) and behind pillars  (see Figure 1D), respectively. Throughout the culture period, media was continuously perfused through the channel. The constant supply of nutrients preserved the viability of the tissue, maintaining spermatogenesis for six months. Goldstein et al.  cultured rat testicular cells on a membrane in a two-chamber device, whose compartments were separated by the membrane (See Figure 7E).
The cells then replicated the blood–testis barrier (BTB) de novo and maintained some of the stages of spermatogenesis. They then evaluated the effect of adding four different known testicular toxins: 1,3-dinitrobenzene (DNB), 2-methoxyacetic acid (MAA), bisphenol A (BPA), and lindane. Using an electrode in each chamber, they monitored the TEER (see Figure 5B) to determine how the BTB was affected. Continuously monitoring the resistance showed that some toxins caused transient damage to the barrier, while others caused more permanent damage. Counting the number of meiotic and post-meiotic cells showed the differential effects of the different toxins on the various stages of spermatogenesis, highlighting the
difference in their mechanisms of toxicity.
A number of relevant organoid models have also been developed. Cham et al.  developed testicular organoids using both fresh and cryopreserved porcine testicular cells (see Figure 7C). After forming organoids in microwells, the organoids were moved to an air–liquid interface setup. Agar blocks were submerged in media with their upper surface exposed. The organoids were placed on top of the blocks and cultured for another four weeks. The organoids’ histology showed both tubular and interstitial compartments. They also developed preliminary vascular structures, which consisted of both immature nascent vessels and vascular structures that were developing into relatively mature micro-vessels, such as capillaries, arterioles, or venules.
A perfused co-culture model was introduced by Baert et. al, who co-cultured liver and testicular organoids on a single OoC chip  (see Figure 7D). First the organoids were formed off-chip. They were then transferred to the chip and the effect of cyclophosphamide, a drug to treat cancer that can only harm germ cells after bioactivation in the liver, was tested. Various biological and metabolic markers were monitored in the testicular organoids. The experiments showed that only the co-culture chips were negatively impacted by the drug, while control chips that only contained testicular cells were not affected. Importantly, this effect was only noticeable in cultures with active flow. Under static conditions, there was no noticeable effect.
Our group, AbuMadighem et al. , has recently developed a ToC platform that allows for the development of testicular organoids de novo on-chip (see Figure 8A), as opposed to the above-mentioned models, where the organoids are first formed off-chip. The chip allows for the addition of various biological factors and hormones throughout the culture period. A comparison to traditional static culture in a well-plate showed that there were more meiotic and haploid cells in the culture. Additionally, cell viability was significantly higher. The same chip can also be used for a perfusion culture. A modified de- sign concept containing multiple compartments allows the co-culture of multiple organoid types on a single chip (see Figure 8C). We have found that culturing disassociated cells in a 2.4 mg type 1 rat-tail collagen gel after incubating the channel overnight with a 50 µg collagen solution results in a single organoid in each channel (without the incubation step, multiple organoids are formed). The setup in Figure 8C could thus be used to culture identical organoids, ensuring uniformity in size.
A tubular model was implemented by Robinson et al. . They used a bioprinter with a coaxial printhead to print tubules (see Figure 7B). The tubular shell contained human
testicular cells suspended in an alginate bio-ink surrounding a polyvinyl alcohol (PVA) sacrificial core. These tubes were then cultured for 12 days. In some areas a structure containing elements of the seminiferous tubule structure was formed with germ cells in the interior, and Leydig and peritubular cells outside the tube. Analyzing the gene expression of the cells in the tubes showed that functional genes that correspond to the self-renewal of SSCs, meiosis, and spermiogenesis were upregulated.
In another study that is relevant to perfusion models, although it was not an in vitro model, Fisher  succeeded in actively perfusing fluid through a seminiferous tubule from a rat, to study the fluid produced by the seminiferous tubules. Ringer medium was perfused through the tubule, while simultaneously the trans-epithelial electrical potential was measured (see Figure 7F).
Future Application of Perfusion Strategies to In Vitro Spermatogenesis
We propose that OoC approaches that incorporate perfusion can significantly improve the ability to recapitulate spermatogenesis in vitro. Organoids can be formed on the chip, or off-chip using existing methods. They can then be cultured in chips with active flow. This would increase the supply of nutrients and the removal of metabolites. Increasing nutrition improves cell viability and reduces the incidence of necrosis in organoids . It also allows for denser cultures, which could improve cellular crosstalk . On-chip sensors could also be incorporated to allow dynamic control of the culture conditions. Active flow also allows for co-culture with various other cell types and OoC models of other organs, as implemented by Baert et al.  (see Figure 7D).
The tubular structure of the seminiferous tubules in vivo makes the testis an ideal organ to replicate using the various tubular OoC models described above. Culturing cells on or in the wall of a perfusable tube would be a more physiologically- and geometrically- accurate representation of the tubules structure. It would also allow access to the lumen (see the concept illustrated in Figure 8B). This would allow for evaluation of the BTB in models with a defined and consistent physiological geometry; similar to the studies performed by Goldstein et al. , using the membrane model depicted in Figure 7E.
The ionic composition and the PH of the fluid in the seminiferous tubules differs from the interstitial fluid found outside the tubule. This fluid plays a key role in creat- ing a suitable environment for the sperm to mature . The fluid is secreted by the Sertoli cells, and the environment is regulated and maintained by the BTB and various ion channels (including voltage gated channels) and water channels . However, the exact composition and nature of this fluid has as of yet not been fully characterized It is difficult to study in vivo or in tubule sections due to the difficulty in retrieving the liquid for measurement . If one or more of the various channel ion channels in the Sertoli cells malfunction, this can lead to infertility, indicating the importance of understanding the composition of the tubular fluid. See Rato et al.  for a detailed review of the seminiferous tubule fluid (STF). Tubular models, such as the concept depicted in Figure 8B, and membrane models, such as the one implemented by Goldstein et al.  (see Figure 7E), can provide access to the luminal compartment and would provide an ideal model to study the fluid. Additionally, if full spermatogenesis were achieved, a tubular or membrane model would allow for the continuous harvesting of mature sperm from the lumen of the tubule, without destroying the model.
There is also an electrical potential difference between the tubule’s lumen and the interstitium [102,105]. The role of this potential has not been extensively studied, but it perhaps plays a role in the regulation of the ionic composition of the tubular fluid. Successfully modeling the BTB in either a tubule model or a membrane chip model that incorporates electrodes, would allow for the study of the electrical potential differences across the barrier, in addition, to the barrier’s permeability and TEER. The electrodes could potentially also be used to artificially induce such an electrical potential in an in vitro model. Adding active flow to such a tubular model, or to other BTB models, would further increase its impact. Mechanical forces, such as shear stresses due to fluid flow, can greatly
impact cellular functions, differentiation, morphology, etc. (see Griffith et al. , Kaarj  and Moraes et al.  for in depth reviews). In the seminiferous tubules, there is pressure gradients and peristaltic-like contractions that create pulsatile fluid flow in the seminiferous tubules [106–108]. This flow transports the sperm through the tubules. The contractions are caused by the contraction of the peritubular cells surrounding the tubules and are controlled by a number of mechanisms. Their frequency and intensity vary throughout the tubule, depending on the stages predominantly present in that section. Defects in the genes that affect the contractile proteins in peritubular cells can reduce the contractility of the tubules and cause infertility .
The above-mentioned studies focus on various factors that influence the contractions. However, many physiological flows and forces also influence the cells themselves [6,7], and perhaps the motion and flow influence the spermatogenic process itself, similar to the gut, where the contraction and flow is seemingly only to propel the food forward, however, it has been shown, in on-chip model  (Figure 5A), that the motion and flow has a significant effect on the morphology and function of the gut model. A perfusion model with either a membrane setup or a tubular model with a lumen would provide an ideal platform to further explore the effects of both the flow and the shear stresses that it applies on the cells, as well as the contractile forces. Recapitulating the BTB in such a model would also allow the study of pressure gradients across the barrier, by maintaining different hydrostatic pressures in the channels on either side of the membrane, or in a tubular model, by maintaining a different pressure in the lumen than in the external chamber, similar to the studies performed by Offeddu et al.  and Bachmann et al.  with vasculature. Even in organ culture models, flow can be introduced into the tubules using the techniques demonstrated by Fisher  (see Figure 7F).
In addition to supplying nutrition, the vasculature interacts significantly with the cells in an organ . In the testis, for example, research has shown that undifferentiated spermatogonial stem cells (SSC) are preferentially localized near the blood vessels in the testes and then migrate when they differentiate [110,111]. Therefore, it would be very significant to culture testicular organoids on vascularized chips  perhaps even using testis-specific VECs. This can be implemented by lining the side channels of a chip with VECs, or by incorporating in the gel of the cell culture channel, or a neighboring channel, a hollow tube the lined with VECs. Alternatively, VECs can be induced to self- assemble into vasculature, in the same channel as the organoids. A further extension would entail culturing organoids that already contain elements of vasculature, such as those developed by Cham et al.  (see Figure 7C), on a chip that includes vasculature, and inducing anastomosis between the two vasculature networks. This could lead to perfusable and vascularized organoids, as was achieved with tumors by Nashimoto et al.  (see Figure 4F). Various strategies to create such perfusable organoids have been discussed By Zhang et al.  and Daniel et al. .
In conclusion, there are many potential benefits to including active microfluidic flow in in vitro spermatogenesis and ToC models. Improving the current models would deepen the understanding of the complex process of spermatogenesis. Such improvement would also increase the probability of these systems providing a viable option for preserving the fertility of pre-pubertal cancer patients and potentially provide a treatment option for adults with maturation arrest.