Findings

In vitro production (IVP) of embryos is currently performed according to standardized protocols for many species and for both commercial and scientific reasons. Oocytes are matured, fertilized and presumed zygotes are cultured in the laboratory until blastocyst stage. Media used in the laboratory usually aim to simulate the natural milieu as closely as possible, e.g. by using synthetic oviductal fluid (SOF) with or without serum for embryo culture. Different protocols use insulin as a stimulatory factor (well known from cell culture as reported in 1976 [1]) to improve embryo development rates.

We performed a pilot study with the aim of further investigating insulin stability in embryo production systems as there is, according to our knowledge, no clear evidence for the choice of the deviating concentrations in vitro.

The media used in our lab contains cysteine in the TCM199 (M2154, Sigma-Aldrich, Stockholm, Sweden).

We added insulin to the in vitro maturation medium consisting of bicarbonate-buffered TCM199 (M2154) supplemented with 0.68 mM l-glutamine (G8540), 0.5 µg/ml follicle-stimulating hormone (FSH) and 0.1 µg/ml luteinizing hormone (LH) (Stimufol; PARTNAR Animal Health, Port Huron, Canada), 50 µg/ml gentamicin and 0.4 % w/v bovine serum albumin (BSA). Groups of 30–45 bovine cumulus oocyte complexes (COCs) were matured in a separate well containing 500 µl medium supplemented either with 0 (INS0), 0.1 µg/ml (INS0.1) or 10 µg/ml (INS10) bovine insulin (I5500). All COCs were incubated for 22 h (24 h after the end of aspiration) at 38.5 °C under a 5 % O2, 5 % CO2 atmosphere. Insulin concentrations were analysed in three different conditions: a) before maturation when the medium itself had been incubated around 2 h for equilibration of temperature and atmosphere and before any COCs were added (BM); b) after maturation with co incubation of COCs for 22 h (AM) and c) after maturation “empty” (AME), a control containing only media without any COCs incubated for 22 h. The Mercodia bovine insulin ELISA Immunoassay for quantitative determination of bovine insulin in serum or plasma specially optimized for bovine samples, was used (Tables 1, 2). The two different insulin concentrations were used to permit the evaluation of a more extreme dose compared to a dose closer to physiological concentrations. The study size did not allow any advanced statistics, only numerical result assessment. Stable concentrations of insulin were found as the added concentrations (0.1 and 10 µg/ml) related well to the measured concentrations after 22 h of incubation. Summarizing, the results of the pilot study do not support the theory that insulin degradation is a reason for using higher insulin concentrations in in vitro studies. However, no statements about the biological activity can be made based on this quantitative ELISA.

Table 1 Insulin concentration in maturation media measured immediately or after 22 h of incubation (INS10 = 10 µg/ml added insulin)
Full size table
Table 2 Insulin concentration in maturation media measured immediately or after 22 h of incubation (INS0.1 = 0.1 µg/ml added insulin)
Full size table

To put our study in the context of recent research, a comparison to recent studies using insulin was made, summarized in Tables 3 and 4.

Table 3 Insulin concentrations used for in vitro experiments based on published literature
Full size table
Table 4 Insulin concentrations measured in follicular fluid and serum based on published literature
Full size table

In the referred studies, insulin is added during the 22 h of in vitro oocyte maturation [2–6], or added to the culture medium during early embryo development [7–9]. Most concentrations used vary between 5 and 10 µg/ml (Table 3). Different effects have been reported, and both beneficial and no significant effects on development have been shown. Even a change in phenotype, in form of increased cell numbers in the blastocyst, has been reported after insulin exposure during oocyte maturation [4, 10]. This could be a sign of accelerated growth of the embryos deriving from these conditions with potentially harmful consequences.

Several studies have been published measuring insulin concentrations in serum and some measuring insulin in follicular fluid, which generally contains hormones and metabolites in concentrations close to those found in serum [11]. The use of different analyzing methods and units (IU/ml; µg/ml; µM/ml) makes comparisons complicated. Most of these studies report concentrations between 0.1 and 1 ng/ml in the bovine (Table 4). The values indicated in the tables are all transformed to the same units; µg/ml in Table 3, respective ng/ml in Table 4, based on the molecular weight of insulin 180.16 mg/mol; 1 mg insulin equals 25.7 or 1 IU = 0.03891 mg.

In the in vivo situation, the developing follicle and maturing oocyte will have a longer exposure to metabolic substances such as insulin. The shortened exposure time might be one factor that could explain that in vitro systems in research often are performed with higher doses than those found physiologically. Another explanation is that in general, in vitro systems have to be pushed with higher concentrations in order to eliminate noise of other influences during the investigated period.

Moreover, the use of supra-physiological doses could be justified by the fact that insulin has been reported to be quite unstable in vitro, due to media containing cysteine [12]. That study demonstrated that a minimum concentration of 1 µg/ml was necessary to stimulate cells in culture resulting in a maximal stimulation 1 h post exposure. Insulin action thus seems to be immediate on the cell. On the other hand, Spicer and Echternkamp [11] report an effective dose 50 (ED50) between 0.01 and 0.2 µg/ml (depending on the effect, as e.g. granulosa cell mitogenesis or estradiol production) in ovarian cell culture, which is closer to the physiological concentrations. These authors used Dulbecco’s Modified Eagle Medium and Ham’s F12 (DMEM/F-12) medium which also contains cysteine.

Knowing that conditions occurring with hyperinsulinemia in vivo (metabolic syndrome, obesity or diabetes) impair fertility, the use of insulin in non-physiological doses should be used cautiously.

Insulin is used in IVP as its stimulatory effects on growth and proliferation have been seen as beneficial for the embryo development while possible negative consequences for the epigenetic regulation and metabolic programming have been largely ignored. Research models should aim to work with concentrations as close as possible to the physiological conditions. Our pilot study does not support insulin degradation as a cause for using 100–1000-fold higher insulin concentrations in in vitro systems than can be found in vivo. Even if no statement about insulin activity can be made, the molecular structure of insulin seems to remain stable—as detectable by ELISA—even in media containing cysteine as in our maturation media based on TCM199. The gap between added and measurable concentrations in the INS10 groups (Table 1) could be explained by the fact that the samples needed to be diluted for the ELISA to remain in the measurement window and this could lead to less accuracy especially as the sample size is low. As the concentrations are more stable in the group closer to physiological concentrations (INS0.1), a general inactivation of insulin does not seem plausible.

However, different reaction and activity profiles in vivo and in vitro are plausible and need to be considered, as in vitro models often aim to investigate consequences with new methods and expensive equipment where an insufficient concentration of the tested factor could cause problems to evaluate the data. More studies focusing on the molecular response of embryos to insulin should be aimed to obtain a better understanding of exposure to insulin during early embryonic development. Unnecessary high doses of insulin should be avoided as the biological consequences of an uncritical use of supra-physiological doses are not clearly known.