Bone marrow from wild-type mice was differentiated with M-CSF to generate macrophages. These macrophages showed an elongated morphology and numerous cytoplasmic vacuoles that we determined were macropinosomes by time-lapse, phase-contrast microscopy (, Video S1). To eliminate the possibility that the LDL receptor mediates LDL uptake and cholesterol accumulation, we also generated macrophages from LDLR−/− mice. Wild-type macrophages were morphologically indistinguishable (including the presence of macropinosomes) from LDLR−/− macrophages (data not shown). The M-CSF receptor tyrosine kinase inhibitor GW2580 inhibited macropinocytosis and caused a loss of macropinosomes from the macrophages (, Video S2). Likewise, withdrawing M-CSF also inhibited macropinocytosis and caused a loss of macropinosomes (Videos S3 and S4), consistent with a previous report showing M-CSF is necessary for macropinocytosis in these macrophages 
. Wild-type macrophages incubated 24 h with 1 mg/ml LDL and GW2580 (5 µM) showed a 46±4% reduction in net cholesterol accumulation compared with macrophages not treated with GW2580 (), thus implicating macropinocytosis in mediating about one-half of macrophage cholesterol accumulation. To determine if GW2580 was active for our experiments, we examined the effect of GW2580 on the M-CSF receptor by Western blotting (Figure S1
). Tyrosine phosphorylation of the M-CSF receptor was similar for vehicle-treated and GW2580-treated macrophages. However, GW2580-treated lysates showed more M-CSF receptor compared with vehicle-treated lysates, indicating that GW2580 was active in some way in our experiments. We also incubated macrophages with LDL in the presence or absence of M-CSF to determine if M-CSF is a mediator of macrophage accumulation of LDL-derived cholesterol. Similar to GW2580, withdrawal of M-CSF from macrophages resulted in a decreased accumulation of LDL-derived cholesterol compared with macrophages incubated with M-CSF (Figure S2
), indicating that M-CSF is a mediator of macrophage accumulation of LDL-derived cholesterol.
Macropinosome formation is M-CSF dependent.
We next incubated LDLR−/− M-CSF-differentiated macrophages with increasing concentrations of LDL. Macrophages showed progressive accumulation of cholesterol (Figure S3A
). The total cholesterol levels increased from 44±2 nmol/mg protein (13±4% esterified) without LDL addition to 518±9 nmol/mg protein (74±1% esterified) when macrophages were incubated with 4 mg/ml LDL. Staining of these cells with Oil Red O to detect neutral lipid revealed massive accumulation of lipid droplets in LDL-treated macrophages while no staining was observed in the absence of LDL (Figure S3
, B and C). To determine whether macrophages incubated with LDL down-regulate cholesterol accumulation with time, we incubated macrophages with 1 mg/ml LDL up to 24 h and assessed cholesterol accumulation (). Macrophages showed a time-dependent increase in cholesterol accumulation showing that down-regulation of cholesterol accumulation did not occur within the incubation period examined. The total cholesterol level increased from 41±4 nmol/mg protein (13±0% esterified) at 0 h to 267±19 nmol/mg protein (72±2% esterified) at 24 h. These experiments show that M-CSF-differentiated macrophages accumulate LDL-derived cholesterol to form foam cells.
Macrophage uptake of 125I-LDL is non-saturable.
Because previous reports from our laboratory show that fluid-phase pinocytosis mediates LDL accumulation in human macrophages 
, we tested whether this also occurs for murine M-CSF-differentiated macrophages. When LDLR−/− macrophages were incubated with increasing concentrations of 125
I-LDL (), a concentration-dependent, non-saturable uptake of native 125
I-LDL was observed, consistent with fluid-phase pinocytosis rather than receptor-mediated endocytosis as mediating uptake of the LDL. To confirm that macrophage fluid-phase pinocytosis mediated LDL uptake, wild-type macrophages were incubated with 125
I-LDL alone or with 125
I-LDL and a 20-fold excess of unlabeled LDL. Macrophage uptake of 125
I-LDL was unaffected by the 20-fold excess of unlabeled LDL, confirming that fluid-phase pinocytosis mediated uptake of the 125
I-LDL (). As a control, macrophages were also incubated with 125
I-acetylated LDL (AcLDL) alone or with a 20-fold excess of unlabeled AcLDL. A 70% reduction of macrophage 125
I-AcLDL uptake was observed when macrophages were incubated with 125
I-AcLDL and a 20-fold excess of unlabeled Ac-LDL compared with macrophages incubated with 125
I-AcLDL alone. This finding of competitive inhibition was consistent with the known scavenger receptor-mediated uptake of this modified form of LDL ().
Fluid-phase pinocytosis mediates LDL uptake.
The scavenger receptors CD36 and SRA are believed to be necessary for macrophage uptake of LDL 
. Our experiments strongly suggest that these receptors are not necessary for macrophage uptake of LDL. To directly determine if CD36 and SRA mediate LDL uptake, macrophages from wild-type, CD36-null (CD36−/−), and SRA-null (SRA−/−) mice were generated and then incubated with 25 µg/ml native 125
I-LDL, or as a control 25 µg/ml 125
I-AcLDL, a modified lipoprotein known to bind to CD36 and SRA 
. CD36−/− and SRA−/− macrophages showed similar levels of 125
I-LDL uptake compared with wild-type macrophages (). We also observed similar levels of 125
I-LDL uptake for wild-type, CD36−/−, and SRA−/− macrophages incubated with 200 µg/ml 125
I-LDL (Figure S4
). As expected, 125
I-AcLDL uptake was significantly lower for CD36−/− or SRA−/− macrophages compared with wild-type macrophages (). These results conclusively demonstrate that the scavenger receptors CD36 and SRA are not necessary for macrophage uptake of LDL.
Because previous studies suggest that the PI3K family of kinases function in mediating fluid-phase macropinocytosis 
, we tested whether this family of kinases is involved in LDL-derived cholesterol accumulation in murine M-CSF-differentiated macrophages. An approximate 50% decrease in net cholesterol accumulation was observed for LDLR−/− macrophages incubated with LDL and treated with the pan-PI3K inhibitors, LY294002 or wortmannin (). To test whether inhibition of cholesterol accumulation was due to inhibition of LDL uptake, we assessed 125
I-LDL uptake by LDLR−/− macrophages. An approximate 50% inhibition of LDL uptake was observed for LDLR−/− macrophages treated with LY294002 or wortmannin (), and thus accounted for the inhibition of net cholesterol accumulation.
The effect of PI3K inhibitors and cytochalasin D on LDL uptake and net cholesterol accumulation in LDLR −/− macrophages.
Fluid-phase pinocytosis mediated by macropinocytosis is an actin-dependent process. To examine whether LDLR−/− macrophage uptake of LDL is dependent on actin polymerization, we assessed net cholesterol accumulation and LDL uptake in the presence of the actin polymerization inhibitor, cytochalasin D (). A 69% inhibition of macrophage net cholesterol accumulation was observed (). Similarly, a 65% inhibition in 125
I-LDL uptake was observed, indicating that inhibition of net cholesterol accumulation was due to inhibition of macrophage uptake of LDL. Macropinocytosis assessed by phase-contrast microscopy was inhibited by LY294002, wortmannin, and cytochalasin D (Figure S5
, A–C and data not shown), suggesting that these inhibitors affect cholesterol accumulation in the LDLR−/− cells by regulating fluid-phase macropinocytosis of native LDL.
Similar to LDLR−/− macrophages, wild-type macrophages treated with LY294002, wortmannin, or cytochalasin D showed a 54%, 59% and 64% inhibition, respectively, of net cholesterol accumulation (). These drugs inhibited macrophage macropinocytosis in these wild-type macrophages (, A–D; Videos S5, S6, S7, S8) indicating that fluid-phase macropinocytosis of LDL was responsible for about one-half of the macrophage fluid-phase uptake of LDL, similar to what was observed for LDLR−/− macrophages. Because wild-type and LDLR−/− macrophages showed similar inhibition of LDL uptake and LDL-derived cholesterol accumulation, these results show that the two types of macrophages are phenotypically similar with respect to macropinocytic uptake of LDL. Consistent with this conclusion, net cholesterol accumulation during a 24 h incubation with 1 mg/ml LDL for wild-type and LDLR−/− macrophages was similar, 136±1 nmol/mg cell protein and 143±1 nmol/mg cell protein, respectively ().
The effect of PI3K inhibitors and cytochalasin D on net cholesterol accumulation in wild-type macrophages.
Macropinocytosis is inhibited by PI3K and actin polymerization inhibitors.
Wild-type and LDL−/− macrophages incubated with LDL accumulate similar levels of cholesterol.
Having observed that PI3K inhibition showed significant inhibition of LDL uptake, net cholesterol accumulation, and macropinocytosis, we examined the function of individual PI3K isoforms with respect to macropinocytosis and macrophage cholesterol accumulation. We assessed LDL-derived cholesterol accumulation by PI3K kinase-dead KI macrophages for the class I PI3K isoforms beta, delta, gamma, or by wild-type mice treated with class I PI3K isoform-specific small molecule inhibitors. No morphological differences including the presence of macropinosomes were observed comparing wild-type and the PI3K kinase-dead KI macrophages (data not shown). All PI3K kinase-dead KI macrophages showed an elongated morphology and macropinosomes, similar to wild-type macrophages. No significant difference in LDL-derived net cholesterol accumulation was observed comparing macrophages derived from wild-type, PI3K gamma kinase-dead KI, and PI3K delta kinase-dead KI mice (). Surprisingly, an increase in net cholesterol accumulation was observed for PI3K beta kinase-dead KI macrophages incubated with LDL compared with wild-type macrophages (). LDL-derived net cholesterol accumulation was also assessed for wild-type macrophages incubated with LDL without or with inhibitors of PI3K alpha, beta, gamma or delta isoforms. Macrophage net cholesterol accumulation and macropinocytosis were unaffected by these PI3K isoform-specific inhibitors (). These results suggest that a PI3K isoform other than class I PI3K mediates murine M-CSF macrophage macropinocytosis of LDL.
LDL-derived cholesterol accumulation occurs independently of class I PI3K isoforms.
The effect of possible inhibitors of fluid-phase pinocytosis on LDL uptake and net cholesterol accumulation in wild-type M-CSF-differentiated macrophages.
Because the molecular components mediating macropinocytosis may vary from one cell type to another 
, on the basis of published literature we tested several potential signaling molecules that may modulate murine M-CSF-differentiated macrophage macropinocytosis of LDL (). To examine the potential function of small GTPases and associated kinases, we monitored net cholesterol accumulation in the presence of LDL and Rac1 or Rho-associated kinase inhibitors (NSC23766 and Y-27632, respectively). Neither of these inhibitors significantly affected net cholesterol accumulation (). Consistent with the lack of effect of these GTPase inhibitors on macrophage cholesterol accumulation, wild-type macrophage macropinocytosis was also unaffected by the presence of these inhibitors (; , A and D; Videos S9, S12, and S14). In contrast, the dynamin inhibitor, dynasore, inhibited net cholesterol accumulation and 125
I-LDL uptake approximately 70% () and also macropinocytosis ( and Video S10). A dynamin specific peptide inhibitor also inhibited macropinocytosis (Videos S15 and S16) and LDL-derived cholesterol accumulation (), confirming that dynamin is a mediator of macropinocytosis. Pak1 kinase and Src-family kinases have been shown to modulate fluid-phase pinocytosis for various cells 
, but inhibitors of these kinases did not alter macrophage net cholesterol accumulation () or macropinocytosis (data not shown).
Macrophage macropinosome formation is dependent on dynamin and vacuolar type H(+)-ATPase.
As Rac1 and Src have been shown to mediate fluid-phase macropinocytosis 
, we sought to confirm that the inhibitors used to target these molecules were active. Untreated macrophages showed GTP-bound Rac1 (i.e., activated Rac1), whereas macrophages treated with the Rac1 inhibitor NSC23766 showed almost no detectable GTP-bound Rac1 as determined by Western blot, confirming the effectiveness of this Rac1 inhibitor (Figure S6A
). Similarly, untreated macrophages contained tyrosine phosphorylated Src-family kinase (i.e., activated Src-family kinase), whereas macrophages treated with the Src-family kinase inhibitor PP2 showed no Src-family kinase tyrosine phosphorylation, confirming the effectiveness of this Src-family kinase inhibitor (Figure S6B
). These results confirm that the Rac1 inhibitor NSC23766 and the Src-family kinase inhibitor PP2 are active in macrophages, and confirm that Rac1 and Src-family kinases do not mediate macrophage fluid-phase macropinocytosis of LDL.
To assess the function of the structural component of microtubules, tubulin, in macrophage LDL uptake and cholesterol accumulation, macrophages were incubated with LDL without or with nocodazole, a microtubule inhibitor. An approximate 20% reduction of net cholesterol accumulation and 60% reduction of 125I-LDL uptake was observed in the presence of nocodazole compared with untreated macrophages (). Nocodazole also seemed to decrease macropinocytosis somewhat, but this was difficult to assess because nocodazole caused the macrophages to narrow (Video S13). Bafilomycin A1, a vacuolar H+-ATPase inhibitor, inhibited net cholesterol accumulation approximately 55% and 125I-LDL uptake approximately 65% for wild-type macrophages (). Macrophage macropinocytosis was also completely inhibited by bafilomycin A1 compared with untreated macrophages (; Video S11).