|
Agent and cell-type specificity in the induction of insulin resistance by
HIV protease inhibitors
|
|
|
Ronit Ben-Romano a; Assaf Rudich d; Dora Torok d; Sharon Vanounou a; Klaris
Riesenberg b; Francisc Schlaeffer b; Amira Klip d; Nava Bashan a,c
Objective:
To test agent and cell-type specificity in insulin resistance induced by
prolonged exposure to HIV protease inhibitors (HPI), and to assess its
relation to the direct, short-term inhibition of insulin-stimulated glucose
uptake.
Methods:
Following prolonged (18 h) and short (5-10 min) exposure to HPI,
insulin-stimulated glucose transport, protein kinase B (PKB) phosphorylation,
and GLUT4 translocation were evaluated in 3T3-L1 adipocytes, fibroblasts, L6
myotubes, and L6 cells overexpressing a myc tag on the first exofacial loop
of GLUT4 or GLUT1.
Results:
Prolonged exposure of 3T3-L1 adipocytes to nelfinavir, but not to indinavir
or saquinavir, resulted in increased basal lipolysis but decreased
insulin-stimulated glucose transport and PKB phosphorylation. In addition,
impaired insulin-stimulated glucose uptake and PKB phosphorylation were also
observed in the skeletal muscle cell line L6, and in 3T3-L1 fibroblasts.
Interestingly, this coincided with increased basal glucose uptake as well as
with elevated total-membrane glucose transporter GLUT1 protein content. In
contrast to these unique effects of nelfinavir, the mere presence of any of
the agents in the 5 min transport assay inhibited insulin-stimulated
glucose-uptake activity. This appeared to be caused by direct and specific
interaction of the drugs with GLUT4 fully assembled at the plasma membrane,
since insulin-stimulated cell-surface exposure of an exofacial myc epitope on
GLUT4 was normal.
Conclusions:
Independent mechanisms for HPI-induced insulin resistance exist: prolonged
exposure to nelfinavir interferes with insulin signaling and alters cellular
metabolism of adipocytes and muscle cells, whereas a direct inhibitory effect
on insulin-stimulated glucose uptake may occurs through specific interaction
of HPI with GLUT4.
Adipose tissue is a key target for HPI in inducing insulin resistance, as the
metabolic syndrome described in patients taking HAART consists of gross
alterations in fat tissue distribution and dyslipidemia, including elevated
circulating FFA and glycerol levels [8]. Adipose tissue is now believed to
play a major role in the pathogenesis of systemic insulin resistance. In
particular, FFA and various adipose-derived peptides (adipokines) have been
identified as modulators of
whole-body insulin sensitivity [43-46]. The systemic insulin resistance seen
in patients taking HAART also involves impaired insulin responsiveness in
skeletal muscle and liver [12]. Consequently, whether these are secondary to
the alterations induced by HPI in adipose tissue or represent direct
effect(s) of HPI in different tissues is of major interest.
The present study investigated the agent and cell-type specificity of HPI in
inducing insulin resistance, focusing on two potential cellular mechanisms.
The data presented demonstrate that, while nelfinavir, indinavir and
saquinavir can directly inhibit glucose uptake when present acutely following
insulin stimulation in cells expressing GLUT4, they manifest differently when
used in a prolonged incubation period. Under the prolonged conditions used,
only nelfinavir elevated basal lipolysis, increased basal glucose uptake
associated with elevated GLUT1 expression, and impaired insulin-stimulated
glucose uptake and PKB phosphorylation. These effects of prolonged exposure
to nelfinavir could be observed in differentiated 3T3-L1 adipocytes, 3T3-L1
pre-adipocytes and in L6 myotubes. Therefore, HPI at the same concentrations
that cause insulin resistance and metabolic alterations in adipocytes, also
affects non-adipocyte cells including skeletal muscle cells. It is,
therefore, conceivable that the reported HPI-induced insulin resistance of
skeletal muscle and liver [12] may represent a combination of direct effects
of HPI in these tissues and actions secondary to adipose tissue alterations.
At the cellular level, the mechanism for HPI-induced insulin resistance and
lipodystrophy appears to be complex. The current literature suggests the
existence of at least three mechanisms: (i) direct inhibition of GLUT4
activity [32, 47]. (ii) activation of basal lipolysis [34, 36, 48] and
inhibition of insulin signals towards GLUT4 translocation [34], and (iii)
interference with pre-adipocyte to adipocyte differentiation [27-30]. The
three mechanisms vary for each agent and time of exposure to the drug. The
direct inhibition of GLUT4 activity is common to all HPI tested ([32] and
Figure 2, Figure 3) and occurs within minutes, even when the agents are
present after insulin stimulation and only in the transport solution. This
appears to be a direct but reversible interaction between the HPI and the
glucose transporters present on
the plasma membrane, since this effect is eliminated upon removal of the
agents. Our data are consistent with the notion that the GLUT4 isoform is
preferentially vulnerable to this acute inhibition by HPI, since this effect
was observed only in GLUT4-expressing cells [3T3-L1 adipocytes, wild-type L6
myotubes (Fig. 2) and L6-GLUT4myc (Fig. 3a)]. In contrast, cells lacking
GLUT4 and expressing GLUT1myc were resistant to acute inhibition of
glucose-uptake activity
by HPI (Fig. 3b). Moreover, by assessing GLUT4 translocation through the
extracellular exposure of the myc epitope in non-permeabilized cells, it is
possible to conclude that the acute effect of HPI does not interfere with the
normal assembly and full insertion of the transporters at the plasma
membrane. Rather, it is likely to represent a direct interaction between
GLUT4 epitopes and the HPI. These results suggest that, in vivo, if the
effective tissue HPI concentrations are sufficient, GLUT4-expressing cells
like adipocytes [32], and skeletal muscle cells (Figure 2, Figure 3 and [49])
may be vulnerable to this direct effect of HPI.
When exposure to nelfinavir was extended to 18 h incubation, 3T3-L1
adipocytes exhibited increased basal lipolysis and elevated basal but reduced
insulin-stimulated glucose uptake [34]. These effects, associated with
enhanced total membrane GLUT1 expression (Fig. 4) and impaired capacity of
insulin to stimulate PKB phosphorylation (Fig. 5), are likely independent of
the direct inhibitory effect that HPI drugs exhibit on glucose uptake, for
several reasons. First, the fall in insulin-stimulated glucose uptake
following prolonged nelfinavir treatment is observed in the absence of the
agent in the transport solution. Second, this lessened insulin action is not
restricted to GLUT4-expressing cells, since 3T3-L1 pre-adipocytes and
L6-GLUT1myc myoblasts also exhibit comparable effects (Figure 4, Figure 5).
Third, while prolonged nelfinavir impaired insulin-stimulated PKB
phosphorylation (Figure 1, Figure 5) and GLUT4 translocation [34], the
short-term effect of HPI did not interfere with PKB phosphorylation [32] nor
with GLUT4 translocation ([32] and Fig. 3c).
Since at least the direct inhibition of GLUT4 and the long-term effect of
nelfinavir may represent independent cellular mechanisms for insulin
resistance, different candidate molecular mediators may be speculated to be
involved. While the molecular basis for the direct interaction between HPI
and GLUT4 has yet to be identified, several hypotheses may be
raised to explain the alterations induced by prolonged exposure to
nelfinavir. Increased expression of proteins like SH2-domain-containing
5-phosphatase and/or protein phosphatase 2A, potentially through interference
with their normal degradation, could result in altered insulin signaling
compatible with the effect of prolonged nelfinavir: impaired PKB activation
in the face of normal phosphatidylinositol 3-kinase activation ([34] and
unpublished data). Increased cellular cyclic AMP levels would explain
elevated lipolysis [50], impaired PKB activation by insulin [51] and
increased GLUT1 expression [52, 53]. Finally, HPI-induced inhibition of
adipocyte differentiation may involve yet other proteins, such as sterol
regulatory element binding protein 1 [29]. It is tempting to speculate that
HPI drugs inhibit a mammalian protein degradation system resulting in
increased expression of proteins, which, in turn, affects lipolysis, insulin
action and differentiation. However, so far, such a mechanism has only been
shown for the degradation of
apolipoprotein B, potentially explaining HPI-induced dyslipidemia [54].
In conclusion, HPI drugs have the potential to cause insulin resistance and
metabolic alterations in various cell types representing targets of insulin
action, through diverse and possibly independent mechanisms. Understanding
these effects at the cellular and molecular level will be essential for both
the design of new HPI with a milder side-effect profile, and the planning of
therapeutic strategies to combat the unwanted effects of HAART.
|
|
|
|
|
|
|