Home
Bank voles as potential new animal model for Type 1, and 2, diabetes.
Background for the discovery.
Polydipsia (excessive drinking) among captive rats was already described by John L. Falk, (1969) but Grethe Sorensen and Axel Randrup (1986) were the first to describe this polydipsia among Danish Bank voles although without making a connection to Diabetes. Nine years later, I (Bryan Schønecker) had started on my thesis work at the Institute of Biology, University of Copenhagen, which started out as an attempt to see if stereotypies were heritable among bank voles (see here for bank vole stereotypies and here for a more general walk through the subject of stereotypies of mammals). At the time it was a widespread belief, at least at my University and probably everywhere else, that polydipsia among Bank voles were a sort of behaviour, maybe stereotypic in nature and maybe related to the phenomenon of scheduled induced polydipsia (SIP), but certainly not simply a disease. However, quite soon I started to note certain peculiarities with these polydipsic Bank voles. They had a markedly increased mortality, tumour-like swellings around eyes, the jaw/neck region and in the armpits (could therefore be related to the lymphatic system); showed weight-loss, tousled fur and a pronounced tendency to walk on the heals after longer time of showing polydipsia. Based on these observations I became more and more sure that this polydipsia among Bank voles were not a part of some kind of behavioural program, but then again, I had not started to speculate what could alternatively be wrong. Presented with my observations it did not take the then graduate student Irene Vejgaard Sørensen more than a couple of seconds to formulate the hypothesis that the Bank voles "obviously were suffering from Diabetes mellitus!".
This diagnose explained my observations perfectly and after that it was a simple matter to confirm the diagnose by measuring strongly elevated blood- and urine glucose values, measurements I made together with a second graduate student, Tonny Freimanis. These elevated urine glucose values were later published in Schoenecker et al., (2000), which described the development of stereotypies and polydipsia among Bank voles along with the hypothesis that polydipsic Bank voles were indeed suffering from Diabetes mellitus. Furthermore the paper provided the observation that stereotypic behaviour was correlated with both an almost 50% reduction of diabetes among females and a delay in the onset of diabetes for those females that later became diabetic. The extreme blood glucose values we measured at the same time were published in Freimanis et al. (2003).
Tonny Freimanis made a thesis work focussing on how different types of stress would influence later development of stereotypies and polydipsia among Bank voles. It was first at an advanced stage of his work that we documented that polydipsia was a very reliably correlate of diabetes among these Bank voles. Subsequent analyses of his data showed that one specific type of stress, implemented once a week the first three weeks after birth seemed to promote the development of diabetes while the same type of stress, performed each day for the first three weeks, had the opposite effect (Freimanis et al. (2003). It therefore seem that the fraction of diabetic Bank voles can be considerably manipulated by using early experienced stress as tool. Exactly what types of stress is most effective for the purpose; which timing duration and frequency is the best is all variables which has only been superficially investigated until now.
As seen in Table 1 (below) roughly twice as many males as females developed diabetes and what could be seen in Freimanis' experiment was that stress applied once a week for the first three weeks drove the females up to the male frequency (which increased a bit too albeit not significantly). Stress applied once a day for the first three weeks post-natal was instead instrumental in reducing the later occurrences of diabetes dramatically.
In my later work, mainly with Knud Erik Heller and Mogens Bildsøe (both from the University of Copenhagen, Denmark), a joint venture was initiated with Bo Niklasson (part-owner of the Swedish private company Apodemus AB) and Ake Lernmark (University of Washington, USA) during which our initial diagnose of Diabetes mellitus among these Danish Bank voles was confirmed and further characterized to be Type 1 Diabetes (Niklasson et al., 2003a).
During a stay in USA (2001) I further found indications that North American Bank voles (C. gapperi) might develop diabetes in captivity (own unpublished data). Later it turned out that Swedish Bank voles too can develop type 1 diabetes in captivity (Niklasson et al., 2003b). Development of diabetes among captive Bank voles therefore seem to be a rather common phenomenon but due to insufficient information's it is still to soon to more than speculate to what extent Bank voles from different regions differ from each other when it comes to prevalence and seriousness in their development of diabetes.
Results confirming the diagnosis of Type 1 diabetes .
Niklasson et al. (2003a) describe below the results of a study on two groups of Danish Bank voles: Group A (n= 101) was killed within 24 hours from capture. Apart from 4 bank voles, which showed blood glucose values > 200 mg/dl (215-343 mg/dl) the average value of the remaining was 101 mg/dl. The pancreas histology was normal among all the tested voles without mononuclear cell infiltrations and they did not show glucosuria. They did however present a broad range of autoantibodies against GAD65, IA-2, insulin and antibodies against the Ljungan virus. Group B (n= 67) was after the capture kept in small barren cages supplied with a woodcutting bed. Food (standard rodent chow) and water was available ad libitum and after a month in captivity the voles were killed and samples analysed. 22 Bank voles (33%) showed blood glucose values in excess of 200 mg/dl (211-540 mg/dl) and of these, 18 Bank voles were positive for ketones bodies, polydipsia or both. I saw signs of hyperlipidemia (white glisten serum) and measured glucosuria and ketonuria in many of these samples. Pancreas was normal in the remaining 45 Bank voles but all the diabetic showed a significant loss of the insulin producing cells, which instead were replaced by prominent vacuolization. Infiltration with mononuclear cells was only seen among a few of the diabetic Bank voles and not among any of the 45 non-diabetics'. Both 87-012 and 145SL Ljungan virus antisera (dilutions 1:4000 or higher) immunostained islets in diabetic voles but not in non-diabetics. Fifteen non-diabetic Bank voles were later inoculated with 145SL Ljungan virus (n inoculated = 10; Control = 5) and were killed after six weeks. All showed normal blood glucose but different degrees of lytic lesions in islets were seen in a [not specified] numbers. The diabetic Group B Bank voles showed significantly higher titres of antibodies against GAD65 and insulin than Group A (killed within 24 hours from capture). This could indicate that Bank voles in the Danish forests is infected with low titres as a consequence and that unknown factors connected to their captivity accelerate the processes leading to higher titres. The diabetic Bank voles had in addition significantly higher levels of anti-Ljungan than the non-diabetic in Group B and the level of anti-Ljungan was correlated to the level of anti-GAD65 (p< 0.03) with a tendency towards a correlation to anti-IA-2 (p= 0.052). Among the non-diabetics in Group B, the level of anti-Ljungan correlated to the levels of anti-GAD65 (p= 0.02), anti-IA-2 (p= 0.005) and anti-insulin (p= 0.004). |
Indications of a Type 2-like diabetes.
Blixt et al. (2007) describes below the results of a study concerning Swedish bank voles. The bank voles (114 males and 105 females) were descendants of a colony founded in 2003 using Swedish voles and were at the time of the study between 5 and 22 weeks old. Different from what had previously been described, the experimental voles were housed in groups of 4-6 voles. Subsequent to a glucose tolerance test (GTT) the voles was classified as either "Glucose intolerant/Diabetic" (GINT/D) or "Normal", dependant on a blood glucose value over, or below 11,1 mM two hours after the GTT. Thereafter, the voles were kept isolated for 48 hours and then killed. Isolated islets were tested for their ability to synthesize and secrete insulin and the ability to metabolize glucose. Roughly 20% of both males and females were classified as GINT/D based on the GTT, and the majority of these voles developed GINT/D between the ages of 5-20 weeks. GINT/D voles showed significantly increased levels of serum insulin compared to "Normal" and tended to have a slightly increased body weight. The islets were balloon-like with a significant reduction in the number of endocrine cells, but the authors note that normal islets were found among the diabetic voles. There was no sign of insulitis. Regarding secretion of insulin, isolated islets from GINT/D males turned out to secrete more insulin than islets from "Normal" males provided the surrounding glucose concentration was low (1.7 mM) where no such difference was observed at higher glucose concentration (16.7 mM). After a week in a glucose containing culture, called RPMI 1640 (11.1 mM) GINT/D males secreted significantly more insulin at the high glucose concentration (16.7 mM) than the "Normal" males. There was at no time differences in insulin secretion between GINT/D females and "Normal" females and the difference there was between the genders concerning the content of insulin in the islets ("Normal" females had significant more insulin than "Normal" males") the first day in culture (day 0) was disappeared after 7 days in culture. The authors speculate that this type of diabetes among [Swedish] bank voles might resemble latent autoimmune diabetes in the adult (LADA) in humans, and "Concerning the evolution of diabetes in the bank vole it is quite likely that the diabetic condition first is characterized more by a type 2 phenotype, and later in elderly animals features of type 1 diabetes may evolve." |
Using Stress as a method to modulate occurence of Type 1 Diabetes.
Freimanis et al., (2003), describe below the results of using post-natal stress within the first three weeks after birth on Danish Bank voles. Blood glucose values: In Schoenecker et al., 2000 we published glucosuria measurements from 20 polydipsic and 20 non-polydipsic Danish Bank voles. This paper by Freimanis et al. included the additional results we got after analyzing their blood samples for glucose. The polydipsic Bank voles showed on average a blood glucose value of 27.7 mM versus the average value of 5.6 mM in the non-polydipsic which confirmed the initial diagnose of Irene Vejgaard Sorensen. Results of applying two types of stress in two different doses: From Table 1 below it is seen that the most efficient method to produce a maximum of diabetic Bank voles is to place them for a short time in lukewarm water once a week. If the objective instead is to reduce the frequency of later diabetics this procedure should be performed once daily for the first three weeks after birth. The procedure is described in more detail in the paper, but shortly it consists of putting some lukewarm water in a one-litre milk container and then let the pups be floating in the water for less than a minute or so. No pup died of this treatment. According to the Table this water-treatment either doubled (3 treatments in total), or halved (21 treatments) the later occurrence of diabetic animals and the key-factor behind this doubling was that the females were pushed up to the levels of the males. There was no gender-related factor to explain the halving of later diabetics. The second method to stress the Bank voles was to separate the pups from their mother for 4 hours - either once a week for three weeks or once daily. The separation was carried out using again a closed one-litre milk carton with woodcutting bed and little holes so they would not lack air.
Figure 1 below depicts the expected (O) and the observed (X) fractions of Type 1 diabetes and the bars indicate the 95% confidence limits around the expected fractions. Model made by Mogens Bildsoe. Figure 1
Figure 1 and Table 1 is both moderated from Freimanis et al., (2003).
|
|||||||||||||||||||||||||||||||||||
Type 1 diabetes in Bank voles maybe reversible after treatment with Pleconaril.
Below is a quote from the publicly available patent application "Treatment of diseases caused by Ljungan virus by using Pleconaril" filed in February 2003 by Bo Niklasson. The quote is from page 9-10 (Examples to support the application): " Diabetic Bank voles, Antiviral compound Bank voles with diabetes were randomised into 2 groups one treated with
Pleconaril and one not receiving any treatment. Treated animals were given 100 mg Pleconaril per kg body weight per day seven days. The drug was administered orally in the
drinking water. All animals were subjected to a glucose tolerance test the day before the
treatment and again 15 days later. Glucose tolerance test was performed by administrating I have to this day (June 2007) not been able to find this experiment publicized in a scientific journal but can the results herein be replicated with a larger sample size of both Bank voles and other species it would certainly be interesting news. |
Prevalence and average age at onset of diabetes.
In practice it has turned out that polydipsia, as defined by a daily water intake in the excess of 21 ml, is a reliably indicator for a diabetic Bank vole. Table two below (modified after Schoenecker et al., 2000) describes the average percentages of Bank voles in my first F1 generation (n Males= 138; n Females= 110) which developed diabetes mellitus (DM) or Diabetes and Stereotypies (DMS) as well as the average age of onset. The number of voles used in this Table is 85 Males and 60 females, which were characterized as either diabetics, stereotypers or both.
| Tabel 2. Average percentages and age (days) of diabetes and stereotypy onset. | |||
| Classification | Males | Females | Ma + Fem |
| % Diabetes (DM) | 30 |
11 |
21 |
| % Diabetes + Stereotypi (DMS) | 9 |
6 |
7 |
| % DM + DMS | 38 |
16 |
29 |
| % Stereotypy (S) + DMS | 35 |
44 |
37 |
| Average age onset: DM | 60 |
74 |
63 |
| Average age onset: S (among DMS ) | 86 |
90 |
87 |
| Average age onset: DM (among DMS) | 62 |
114 |
79 |
From the table over this cohort it can be seen that over twice as many males as females developed diabetes and the same was true even if the diabetics were summed together with the voles that developed both diabetes and stereotypies (p< 0.001 in both cases).
Furthermore it can be seen that diabetes among DMS voles seemed to have a later onset (79 days vs. 63 days; p non-significant), a difference which was particular evident among the females (114 days vs. 74 days; p< 0.05).
Lastly this table can be used to calculate that non-stereotypers seemed considerable more prone to develop diabetes than stereotypers (34% vs. 20%; p< 0.05).
Average daily water intakes for Bank voles.
My thesis ("Stereotypical behaviour among Bank voles", Zoological Institute, University of Copenhagen, 1998) involved 92 (P); 248 (F1) and 270 (F2) Bank voles and in Table 3 below it can be seen that the maximum measures of daily water intake in these two captivity-born generations is quite comparable with the data we published in Schoenecker et al., 2000, a paper which only dealt with the P- and F1generation.
A non-diabetic/non-stereotypic vole is in the following designated "W" (wildtype); a diabetic "DM"; a stereotypic "S" and a diabetic/stereotypic "DMS".
| Tabel 3. Maximum daily water intake (ml) | ||||
| Classification | n Males/Females | Males | Females | p-value |
| W (F1) | 52/51 | 9.8 |
9.0 |
N.S. |
| W (F1+F2) | 87/106 | 12.3 |
11.9 |
N.S. |
| DM (F1) | 41/12 | 67 |
60.8 |
N.S. |
| DM (F1+F2) | 103/39 | 68.7 |
59.1 |
p= 0.0013 |
| S (F1) | 32/42 | 9.2 |
10.3* |
N.S. |
| S (F1+F2) | 65/74 | 12.4 |
15.1* |
p= 0.0003 |
From Table 3 it can be seen that a diabetic male consumes a bit more water than a diabetic female (Males weigh a bit more too) and this water intake corresponds roughly to three times their body weight.
Using only data from F1 revealed no gender differences at all, only that S females drank more water than W females (p< 0.05). Combining data from F1 and F2 resulted in a clear gender difference among both the diabetics (males drank the most; p= 0.0013) and the stereotypers (females drank more; p= 0.0003). Again the stereotyping females from F1+F2 drank more than the non-stereotyping females (p= 0.0001) but the reason that performing in stereotypies increase only the female consumption and not the male- is probably interesting, but at present unresolved.
Easy method to screen for diabetic Bank voles.
The easiest method is simply to keep the bank voles in transparent cages and take a glance through the bottom while changing the cages with clean ones. If there is a beginning, or big, reddish/brown spot with a greasy appearance (bacteria) you have a polydipsic, and hence diabetic bank vole. It's easy to measure the water consumption each time you provide the Bank vole with a fresh bottle - just weigh the bottles and most databases has a function to insert a date and time and use such data in a homemade formula to calculate the daily water consumption. Shift the bottles once a week and you will probably have no problems with calcium in the bottlenoses. You can measure the weight too as additional data.
I cannot recommend using orbital bleedings to provide blood samples for glucose measurements. Many of my Bank voles later contracted severe eye problems and there is absolutely no reason to expose the animals for un-necessary sufferings. It is quite feasible to take a blood sample from the saphenous vein in the lower leg instead - a technique I used with no problems at all in Seattle. Read this for a thorough description or read Hem & Solberg (1998).
Summing up:
Subsequent to capture and transport to the place where it is convenient to extract blood samples to measure glucose there are relatively few Danish voles that exhibit hyperglycaemia as opposed to Swedish Bank voles which have elevated blood glucose in rich numbers (Niklasson et al., 2003b; 2006a). Apparently there were nothing wrong with their pancreases in this study so to diagnose them as diabetics in the wild is to stretch the definition of what is required to be a diabetic. It has been suggested that the reason for this hyperglycaemia among Swedish wildcaugh Bank voles should be social stress because of a high population density (Niklasson et al., 2003b; 2006a). However that explanation do not seem likely to me for various reasons.
After a while in captivity, Danish bank voles develop clinical signs of type 1 diabetes in high frequencies. The average age for onset is typically a couple of months for captivity-born males and roughly 0.5 - 1 months later for the females. (Schoenecker et al., 2000). The most blatant signs are a total destruction of the insulin producing islets accompanied with prominent vacuolisations (Niklasson et al., 2003a).
Diabetic Bank voles harbour antibodies against GAD65, IA-2, insulin and Ljungan virus (a virus which have a certain resemblance to GAD65 - Niklasson et al., 2003a). They loose weight and develop various symptoms of diabetes-related diseases after longer times of suffering from diabetes.
Infiltrations of mononuclear cells were only seen among a few Bank voles.
Males normally become diabetic twice as often as females, but females can easily be "stressed-up" to male levels.
Swedish bank voles apparently develop a type 2-like diabetes (Blixt et al., 2007) but surprisingly enough, the author's refrain from discussing the possibility of genetically based differences concerning the diabetes which is evident among Swedish and Danish bank voles, geographically separated as they have been for thousands of years. Natural differences in genetic make-up could be thought to influence the prevailing type of diabetes, the prevalence and age of onset.
The diffuse concept of "Stress" is consequently, as first demonstrated by Freimanis et al. (2003), clearly a factor which can be used to modulate the prevalence of diabetes among captive Bank voles. Using the proper methods it is to be expected that up to 50-60% of Danish Bank voles will develop diabetes mellitus in captivity.
Ljungan virus may be involved as ethiological agent behind bank vole diabetes but so far (May, 2010) only two papers (Niklasson et al. 2003a and 2003b) have specifically addressed the issue. The results associated presence of Ljungan virus antibodies with diabetes in these animals, but any causality has obviously not been proven. In fact (and quoting from Blixt, 2010): "Attempts to determine with accuracy if an individual diabetic bank vole is LV infected or not, have so far not been successful."
As far as I am informed it is still not possible to acquire virus-free Bank voles (Niklasson et al., 2003b; 2006a; 2006b; Blixt et al., 2007) but I assume that can only be a question of demand.
Conclusion:
Based on the available information's Bank voles seem to be a suitable animal model to study the complications of an untreated diabetes. The tendency to develop this disease seems to be easily moderated by early-experienced stress which makes the Bank voles promising to study the causalities behind diabetes. Ljungan virus seems to be an important etiological factor behind Bank vole diabetes but that remains to be proven using a Ljungan virus-free colony.
An interaction between stereotypies and diabetes is evident.
Bank voles are easy maintained in a stable; they are large enough to deliver blood samples and they breed well in captivity so if the possibility comes around then here is an interesting new lean rodent model for type 1 diabetes. Compared to NOD mice and BB rats this model has only been most superficially investigated, which in my opinion is only a further encouragement.
Last updated the 27th May, 2010