Evaluating the effect of embryonic exposure to antiretroviral therapy (ART) on the development of type-2 diabetes mellitus (T2DM) in a zebrafish model.

Abstract

The use of combination antiretroviral therapy (cART) has significantly increased the quality of life and lifespan of people living with HIV (PLWH), however this new lease on life is threatened by a high risk of developing metabolic disorders such hyperglycemia and type 2 diabetes mellitus (T2DM). TLD, a fixed-dose combination antiretroviral drug comprising tenofovir (300 mg), lamivudine (300 mg) and dolutegravir (50 mg) is a first-line treatment regimen for all PLWH in South African and other mid-to-low income countries, as recommended by the World Health Organisation (WHO). In adults, TLD is however increasingly associated with adverse side effects such as hyperglycemia, bodyweight gain and T2DM. The mechanisms underlying these effects are not well understood and indicate the need to develop tools and models to better understand the development of cART-related T2DM. Of critical importance, TLD is also widely used for viral suppression in HIV-positive pregnant women as it crosses the placenta, conferring some level of protection in the developing fetus and thereby reducing the risk of mother-to-child transmission. However, whether in utero exposure to TLD predisposes the embryo to developing metabolic diseases such as T2DM later in life are not known. Studies on the short- and long-term effects of TLD exposure in utero are therefore required. In this context, the zebrafish (Danio rerio) is seen as an ideal animal model for many human diseases and several zebrafish models of T2DM exist. The aim of this study was therefore to utilise zebrafish to increase our understanding of the effects of embryonic TLD exposure on the development of T2DM in adult organisms, with post-natal bodyweight, fasting blood glucose, and the relative qPCR expression of hormones and enzymes central to glucose metabolism as indicator endpoints. In order to establish morphological (bodyweight), glycemic (fasting blood glucose) and relative qPCR genetic changes likely to occur in non-obese and obese diabetic zebrafish, we first developed adult zebrafish models of T2DM using glucose-immersion and overfeeding approaches. The mean bodyweight of the overfed group (0.73 g ± 0.05 g) was significantly higher than that of the control group (0.57 g ± 0.1 g), while there was no significant bodyweight change for the glucose model. The fasting blood glucose levels of the overfed (3.5 ± 0.9 mmol/L) and glucose-immersed (4.0 ± 0.7 mmol/L) zebrafish models were significantly higher than the controls indicating the progression to insulin resistance and T2DM. Both models revealed a lack of insulin and preproinsulin expression in some fish within these treatment groups relative to controls when assessed with qPCR (i.e. no target amplification). The toxicity of TLD and its individual components in vitro (on HEK293 cells) and in vivo (on one-day-old dechorionated embryos), and thereby the appropriate TLD concentration for treatment, were then determined. TLD did not induce toxicity in HEK293 cells even when used at the highest concentration (of 100 μM Tenofovir, 100 μM Lamivudine and 50 μM Dolutegravir), but did slow the rate of cell growth. TLD was also not toxic to zebrafish embryos after five days of exposure, but did slow their development by delaying the formation of the swim bladder and therefore the ability of developing embryos to maintain an upright posture. One-day-old dechorionated zebrafish embryos were then exposed to TLD for 5 days, followed by the evaluation of bodyweight and fasting blood glucose changes in adult zebrafish at 4 months. The bodyweight (0.26 g ± 0.12 g) and fasting blood glucose (3.5 ± 1.92 mmol/L) in response to TLD treatment was higher, but not significantly different to the controls (0.17 g ± 0.04 g and 2.0 ± 0.62 mmol/L) respectively. The relative expression of insulin (zins), preproinsulin (ppins) and phosphoenolpyruvate carboxykinase 1 & 2 (pck1 & pck2) (all referred to as candidate genes) was then assessed using qPCR. The qPCR expression of insulin (Δ𝐶𝑡 2.02 ± 1.0) and preproinsulin (Δ𝐶𝑡 1.87 ± 0.9) of pooled samples at 4 months after TLD treatment was variable but still higher than controls (Δ𝐶𝑡 1.0 ± 0.0). The expression of all the selected candidate genes at 5 days and at 1-month post treatment was not significantly different to controls. In conclusion, to study the implication of embryonic TLD exposure on the development of T2DM in adult organisms, we first established T2DM models in zebrafish in order to understand and establish the changes (physiological and genetic) reflective of a diabetic state. We demonstrate that changes in relative qPCR gene expression in adult zebrafish was variable in both models, with some fish displaying elevated levels while others lost the ability to express insulin and its precursor preproinsulin. Furthermore, our study suggests that exposure to TLD during embryonic development may contribute to weight gain and hyperglycemia (i.e. insulin resistance) later in life, even long after the drug is removed. The genetic effects of TLD were not clearly visible, however it is possible that dietary challenge in the adult fish is required to demonstrate the extent of the effect. These preliminary findings provide insight on the implications of embryonic TLD exposure in the context of the development of metabolic diseases such as T2DM in adults.