TL:DR at bottom
Is insulin resistance due to a graded loss of insulin response at the individual cellular level or does it reflect changes in the fraction of cells responding to insulin?.......
.... is the crux of a new paper(1) concerned with the insulin sensitivity of fat cells. We are all very use to the idea that the insulin sensitivity of a fat cell varies with its size such that large overstuffed fat cells are insulin resistant (IR) and small fat cells are insulin sensitive (IS). So, with the above question in mind, it seems obvious that we should chose the former .... I.E. "its a graded response, varying with the size of the adipocyte"....... HOWEVER, this infact may not be true.
Or atleast, only partially true..
The first caveat to note, is that they are using GLUT4 translocation as the surrogate for the IS of the fat cell.
Systemic insulin responses, such as glucose clearance, represent the integrated GLUT4 translocations of all responding muscle and adipose cells
So the more GLUT4 translocation we get in a fat cell, the more IS it is deemed to be. Previously, the authors had discovered that human adipose cells cultured
in vitro WITHOUT the hosts serum retain the IS status of their host. This would imply that circulating factors in host serum (i.e. hormone levels etc) may not play as important role in the IS of the individual as previously thought, and instead suggests that the IS of a fat cell is an intrinsic property of that cell.
Another interesting point is that, when adipose cells are isolated and tested individually for their IS, there is marked heterogeneity between cells in their responses.(2) That is, each individual adipocyte will display a unique and highly reproducible response. As the authors of (2) say...
These data highlight that the response of a cell population to insulin is underpinned by extensive heterogeneity at the single cell level. This heterogeneity is pre-programmed within each cell and is not the result of intracellular stochastic
Back to (1), they sought to ask the question and test, is the IS status of a host reflected in the IS status of all his individual fat cells. For example, if someone is 50% overall IR, is that because all his individual cells are 50% IR?
changes in adipose function are usually studied at the whole organ or population level and it is naturally assumed that any defect will be due to a similar disruption in all cells(2)
Testing and Data from (1)
The tests performed by (1) were to isolate 8 individual fat cells from the abdominal subcutaneous depot of 19 different subjects, all of whom showed a very wide variation in their whole body IS status.
The cells were incubated in the absence ( basal ) or presence of 0.1 IU/ml insulin (maximal stimulation) and GLUT4 translocation markers were assessed.
The graph is plotted with the most IR subject on the left and the most IS subject on the right. And the graph proceeds from left to right in terms of the IS of the subjects...
Each blue dot represents an individual fat cell from 1 person in the basal state, and the red dot next to it is one of the individual fat cells in the insulin stimulated state. As you can see, as you move from left to right, more and more red dots jump above the dotted line, indicating that as you move from left to right, more and more cells respond to the insulin. Red dots that stayed below the dotted line means they didnt respond to the insulin at all.
Interestingly, the major difference between cells from insulin-resistant and insulin-sensitive is not the individual cell response amplitude, but rather the number of cells that exhibit a 3–4 fold response. Simultaneously, in almost every subject, we observed some cells that do not exhibit any insulin response at all.
What this seems to suggest, is that the difference between IS and IR individuals is not how much the cells respond to insulin, but rather the number of cells that respond. I.E. Insulin resistant people have a much lower number of cells that respond to insulin. I found this very counter-intuitive and novel.
Infact, if you plot all the data on a BeeSwarm style graph...
....you get a graph that looks like a bimodal function. I.E. 2 distinct states of normal distributions. Based on this the authors propose that fat cells exists in 2 distinct states, "basal" and "insulin stimulated".
Thus, GLUT4 translocation in individual adipose cells isolated from human subjects is better characterized as an all-or-none (bimodal) response than a graded response. Our single cell analysis of human adipose cells reveals a bimodal distribution for each of two characteristic properties of GSV, mobility and fusion, affected by insulin. This invariance in the underlying distributions, refractory and insulin-stimulated, with a clearly determinable fraction of cells populating these two states in each subject, is consistent with a threshold or switch-like transformation from the basal to the stimulated state upon addition of insulin
Threshold type triggering systems exists everywhere in biology, and it wouldnt surprise me if something similar exists for how adipose cells respond to insulin. I (speculatively) envisage that each individual adipose cell probably has a threshold insulin responding level, if it sees insulin below that threshold concentration, no response is initiated. However as soon as it sees insulin above the threshold concentration, it immediately enters the "insulin stimulated" state and simultaneously fires its entire GLUT4 cache. This threshold level is probably changeable depending on hormonal inputs into the adipose cell.
Infact there is some evidence that the insulin sensitivity is dependent on the number on insulin receptors on it........ This quote from ref(2) suggests insulin sensitivity of a fat cell is dependent on receptor count.
"previous studies have shown that changes in insulin receptor number can change the sensitivity but not the maximum responsiveness of insulin action(3) ."
Even more intriguing from (1), the authors ruled out that fat cell size was the deciding factor in whether or not a fat cell responded to 0.1 IU/ml insulin, as no correlation was observed between fat cell size and GLUT4 translocation.
Findings from study ref (2).
The authors in ref (2) worked on
groups adipocytes
but performed atleast 1 study involving single adipocytes, (3T3-L1 Adipocytes), but they still found evidence of extensive heterogeneity.
From fig 5a. , the response of the
group of fat cells to different insulin doses was a graded response.
They noted that at the lowest insulin dose, only 30% of cells from the group displayed any detectable response, which is consistent with the idea that each cell has a threshold insulin trigger level that must be passed before it will show any insulin response.
However, somewhat contradictory to ref(1), they found fat cells that responded in both a graded response manner
and a bimodal response...
I emailed the authors of ref(1) about this seeming contradiction, ( because (1) declared fat cells as being bimodal while (2) found evidence for graded responses aswell as bimodal) and specifically asked how could they determine the presence or absence of a graded responses if they only tested with 1 different insulin dose. They have yet to email me back.
Anyway......
Three important conclusions from ref(2) are....
- These data imply that the cellular response to a specific dose of insulin is an intrinsic property of each cell.
- One possibility is that this represents discrete subpopulations of adipocytes that possess intrinsic differences in insulin sensitivity and responsiveness.
- changes in adipose function may rather depict changes in the relative abundance of subpopulations of adipocytes that comprise the organ.
Currently, adipocytes are divided into 3 subpopulations, (white, beige, and brown.) What seems likely is that even amoung white adipocytes they can be divided into even further subpopulations that differ in insulin responsiveness and sensitivity.
Personally, what *I* think is happening is that each individual cell has its own level of epigenetic expression of its gene's ( epigenetic = histone acetylation and DNA methylation ) that together makeup the phenotype of that cell. A large fat cell has increased histone acetylation on PPARgamma, and its possible epigenetic expression of that aswell as other gene's will determine its insulin responsiveness.
Are small fat cells really more insulin sensitive?
Ref(4) found an enhanced number of small fat cells in insulin resistant subjects compared to insulin sensitive ones. This is
not what you would expect if you just bluntly thought small fat cells = insulin sensitive.
Here's the key quote from the abstract...
The real-time PCR results showed two- to threefold lower expression of genes encoding markers of adipose cell differentiation (peroxisome proliferator-activated receptor gamma1 [PPARgamma1], PPARgamma2, GLUT4, adiponectin, sterol receptor element binding protein 1c) in insulin-resistant compared with insulin-sensitive individuals.
Although the authors of ref(4) attribute this to"impaired" adipogenesis, I dont agree. The lowered expression of these adipose gene's *is* what makes the fat cells smaller and probably also more insulin resistant. But this is not a "defect" , but rather it is a reflection of their genetic inheritance.
We know insulin resistance has a genetic component and runs in families. I reckon what is happening is that offspring of insulin resistant parents are simply inheriting pre-adipocyte pools that have lower levels of histone acetylation of those adipogenic gene's in them. (particularly PPARgamma)
Remember from ref(5), the histone acetylation ( H3 anyway ) of an adipocyte is maintained even in the de-differentiated state. This implies the pre-adipocyte
already contains a blueprint for exactly what type of fat cell it should become ( large vs small etc ), if and when it ever gets the signal to turn into a fat cell.
So.... as this normal person starts to eat obesogenic food ( processed carbs ) , morphological changes to the intestine and pancreas takes place, causing hyperinsulinemia and hyperglycemia, which in turn start to activate all his pre-adipocytes and turn them into mature fat cells. These new fat cells start to store their own fat ( thats what fat cells do, calories are irrelevant, exactly like how myocytes are going to build myofibrils even if you keep protein intake low.. ) , and the person overall gets "fatter". No doubt the insulin also causes existing adipocytes to hypertrophy.
Depending on the genetics and epigenetics of the pre-adipocyte pools you inherited from your parents, these pre-adipocytes could turn into small insulin resistant fat cells, in which case youll only gain *some* weight.and as your insulin levels continue to increase, putting increased insulin sensitivity demands on your body, these fat cells cant fulfill that demand for increased insulin sensitivity, and youll been seen as "insulin resistant" and labelled diabetic ..........Or they could turn into large insulin sensitive fat cells, in which case youll get super obese. I also expect numbers pre-adipocytes in pools differs between individuals, which would further contribute to your propensity to get extremely fat.
So when a diabetes researcher harvests fat tissue from an obese insulin resistant subject and observes them under a microscope and sees lots of small fat cells and concludes "defect in adipogenesis!" . I say NO.
I say, your just observing the genetically inherited destiny.of his cells.
TL:DR conclusions
- (1) concluded that fat cells respond to insulin ( GLUT4 translocation ) in an all or nothing manner.
- It appears fat cells have an insulin threshold trigger level, if they see insulin below the threshold, the response is zero, but if they see insulin above that level, they immediately respond.
- contrary to (1), (2) found evidence that the response of individual fat cells to insulin above the threshold can be both graded, and bimodal.
- graded responding fat cells will respond stronger as insulin doses increase, up to maximum dose, while bimodal responding fat cells always give the same response regardless of insulin dose.
- (1) found that some fat cells will not show any response to insulin even at maximal doses.
- the difference between insulin sensitive and insulin resistant individuals appears to be accounted for entirely by the number of fat cells that just stop completely responding to insulin. That is, as you progress from insulin sensitive to insulin resistant, more and more of your fat cells will completely ignore insulin
- this defect is preserved in culture even without host serum, indicating that the response to insulin is an intrinsic property of each individual fat cell.
- (4) found that insulin resistant individuals have a greater proportion of smaller fat cells with reduced expression of adipogenic gene's, countering the idea that smaller fat cells are always more insulin sensitive.
Refs
1. Human Adipose Cells In Vitro Are Either Refractory or Responsive to Insulin, Reflecting Host Metabolic State -
link
2. Novel systems for dynamically assessing insulin action in live cells reveals heterogeneity in the insulin response. -
link
3. Mechanisms of insulin resistance in obesity and noninsulin-dependent (type II) diabetes. -
link
4. Enhanced proportion of small adipose cells in insulin-resistant vs insulin-sensitive obese individuals implicates impaired adipogenesis. -
link
5. Heritability of fat accumulation in white adipocytes. -
link