Cardiopulmonary Bypass Decreases G Protein–Coupled Receptor Kinase Activity and Expression in Human Peripheral Blood Mononuclear Cells

After written informed consent was obtained, fifteen patients were enrolled in this study approved by the Institutional Review Board. The patients underwent surgery involving CPB and were enrolled successively without exclusions for patient condition or type of surgery. Blood samples (15 ml) were collected in EDTA-coated tubes at the following times: before surgery (PRE), 2 h after completion of CPB (POST), and 24 h after completion of CPB (POD1). Samples were transported to the laboratory at room temperature. Patient records were reviewed to gather demographic data, preoperative diagnosis, surgical procedure, length of CPB, length of hospital stay, and postoperative complications.

An 800 μl portion of each sample was centrifuged at 1500 ×g  to obtain plasma used for the IL-6 assay. The remaining blood was used to isolate peripheral blood mononuclear cells over Ficoll-Hypaque Histopaque® (Sigma, St. Louis, MO) density gradients, as described by Böyum. 21Mononuclear cells and plasma were stored at −80°C until used.

Mononuclear cells were thawed, then suspended in 0.5–1.0 ml of ice-cold buffer A (20 mm Tris-HCl [pH 7.4], 2 mm EDTA, containing the protease inhibitors 100 μm PMSF, 10 μg/ml leupeptin, 5 μg/ml aprotinin, 10 μg/ml soybean trypsin inhibitor, 150 μg/ml benzamidine); homogenized using a hand-held tissue disruptor on low speed; and centrifuged at 4°C at 45,000 g  for 15 min. The supernatant, representing the cytosolic fraction, was stored at −80°C until analysis. The membrane pellet was resuspended in 0.5 ml buffer A and stored at −80°C.

The concentration of plasma IL-6 was determined by enzyme-linked immunosorbent assay using a commercially available kit following the manufacturer instructions (Quantikine HS, R&D Systems, Minneapolis, MN).

The GRK activity in the cytosolic fraction was assessed using light-dependent phosphorylation of rhodopsin in urea-washed bovine rod outer segments (ROS), as described previously. 22,23Briefly, 6 μg of cytosolic protein from mononuclear cells and 5 μg of ROS proteins were suspended in a total volume of 50 μl consisting of 20 mm Tris-HCl (pH 7.4), 10 mm MgCl², 2 mm EDTA, and [γ-³²P] adenosine triphosphate (final concentration 100 μm, specific activity 2–4 cpm/fmol). The reaction was allowed to proceed for 5 min at 30°C and was then stopped by adding 50 μl of sample buffer (25 mm Tris-HCl [pH 6.8], 10% glycerol, 8% sodium dodecyl sulfate, 0.0125% Bromophenol Blue). Phosphorylated rhodopsin was separated by electrophoresis on 10% sodium dodecyl sulfate–polyacrylamide gels (SDS-PAGE), and visualized by autoradiography. The rhodopsin band was excised from the dried gels and counted by liquid scintillation spectroscopy. Background phosphorylation was determined by performing phosphorylation of rhodopsin in the absence of mononuclear cell cytosol, and ³²P incorporated into rhodopsin was subtracted from ³²P incorporated into rhodopsin in the presence of mononuclear cell cytosol. Results are expressed as fmol ³²P incorporated into rhodopsin · minute⁻¹·μg cytosolic protein⁻¹, and as a percentage of baseline (PRE) activity.

The expression of GRK2 and GRK6 proteins in the cytosolic and membrane fractions of mononuclear cells was determined by Western blot. Samples containing 30 μg of protein were placed in SDS-sample buffer, incubated for 15 min at 37°C, and subjected to electrophoresis on 10% SDS-PAGE gels. The separated proteins were transferred to PVDF membranes (Millipore, Bedford, MA). The membranes were incubated for 1 h in a 1:200 dilution of primary rabbit polyclonal antibodies to GRK2 or GRK6 (Santa Cruz Biotechnology, Santa Cruz, CA) followed by 1 h incubation in a 1:2000 dilution of antirabbit secondary antibodies conjugated with horseradish peroxidase (Santa Cruz Biotechnology). Immunoreactivity was detected by enhanced chemiluminescence (Pierce, Rockford, IL), and quantified by densitometry using Quantity One software (Bio-Rad Laboratories, Hercules, CA). The signals were determined to be in the linear range for chemiluminescence. The GRK protein expression in POST and POD1 samples is reported as a percentage of the PRE sample for each individual. Purified recombinant GRK2, prepared as described previously, 24was used as standard in GRK2 immunoblots. Recombinant GRK6 in Sf9 membranes was used as standard in GRK6 immunoblots. To express GRK6 in Sf9 cells, human GRK6 cDNA was excised from ATCC clone HFBCB24 (ATCC, Rockville, MD) with EcoRI/Bsp 120I (Apa I) and subcloned into PVL1393 digested with EcoRI/NotI. A recombinant baculovirus containing GRK6 was isolated using the protocol provided with the BaculoGold transfection kit (Pharmingen, San Diego, CA), and it was used to infect Sf9 cells. The growth of Sf9 cells and preparation of Sf9 membranes have been described previously. 25 

Only GRK2 and GRK6 were measured because the commercially available antibody to GRK3 also reacted with purified GRK2, and levels of GRK5 were too low to quantify. Whether our GRK2 antibody cross-reacts with GRK3 could not be assessed because purified GRK3 was not available. The GRK6 was analyzed only in mononuclear cell membranes because it is a membrane-associated enzyme. 19The GRK1 and GRK4 were not examined because these GRKs are only expressed in retinal and testicular tissue, respectively. 19 

The GRK activity and protein expression for each subject are reported as a percentage of PRE levels and expressed as mean ± SD. An exact Wilcoxon signed rank test or paired t  test was used to study change in GRKs and IL-6 values between PRE, POST and POD1 samples; we applied the Bonferroni correction to these changes and a P  value of less than 0.0167 was considered significant for these tests. The relationship between these measured variables were determined using exact tests of Kendall τ b correlations, simple multiple linear regression, and separate mixed models. A P  value of less than 0.05 was considered significant for these tests. Data were analyzed using SAS software (SAS, Cary, NC) and STATXACT (Boston, MA) software.

Table 1presents patient demographic data, surgical procedure, length of CPB, length of postoperative hospital stay and postoperative complications.

Plasma IL-6 increased more than 20-fold after CPB from 2.9 to 62.0 pg/mL (P < 0.0001) and remained increased to 88.6 pg/mL on POD1 (P < 0.0001) (fig. 1).

Figure 2Ashows that rhodopsin phosphorylation, which represents GRK activity, 26,27decreased substantially in the POST sample and remains low in the POD1 sample. Average kinase activity in PRE samples was 7.5 ± 3.5 fmol ³²P incorporated into rhodopsin · minute⁻¹·μg cytosolic protein⁻¹, and it decreased by 39 ± 29% after CPB (P = 0.0014). The CPB-induced decrease in GRK activity was seen in all but one patient, but the recovery was quite variable among patients (fig. 2B).

We determined whether the observed decrease in GRK activity after CPB was due to a decrease in GRK2 protein. Figure 3shows that cytosolic GRK2 expression decreased dramatically after CPB by 90 ± 15% (P = 0.0001), indicating that the decrease in cytosolic GRK activity after CPB could be attributed to a decline in GRK2 protein. The decrease in cytosolic GRK2 was not due to GRK2 translocation to membrane because membrane-associated GRK2 was also significantly decreased after CPB by 78 ± 27% (P = 0.002, data not shown). It should be pointed out that the decrease in GRK2 protein in POST samples (fig. 3B) was much greater than the decrease in GRK activity (fig. 2B). The disparity between measured GRK activity in the cytosol and the expression of GRK2 protein in the cytosol indicates that other kinases contribute to the phosphorylation of rhodopsin.

As with GRK activity, the recovery of GRK2 on POD1 differed considerably between patients (fig. 3B).

To determine whether the decrease in GRK2 after CPB was protein-specific, we examined the effect of CPB on the G proteins G^sαand G^iα(proteins which, like GRKs, are part of the GPCR signaling network). 19,28As Figure 4shows, the effect of CPB on G^sαand G^iαproteins does not parallel the effect on GRK2. Thus the effect of CPB on GRK2 decline is GRK2 specific and is not due to a CPB-induced global degradation of proteins. Our finding that CPB had a variable effect on G^sαand G^iαproteins in mononuclear cells is consistent with a recent report by Tittelbach et al. which found that G^sαand G^iαprotein expression in peripheral blood mononuclear cells was increased in some patients after CPB, and decreased in other patients. 28 

Figure 5shows that membrane-bound GRK6 decreases after CPB and the average decrease in GRK6 is 65 ± 43% from PRE levels (P = 0.0015). As with GRK2 and GRK activity, the recovery of GRK6 on POD1 varied considerably between patients (fig. 5B).

A negative correlation between cytosolic GRK activity and plasma IL-6 was detected when a mixed model analysis was performed (P = 0.0017) and a negative correlation was also found between GRK6 expression and IL-6 using simple linear regression (P = 0.0181). However, we did not find a statistically significant correlation between changes in GRK2 expression and plasma IL-6, presumably due to our small sample size.

In this study, we provide the first evidence that cardiac surgery involving CPB decreases GRK activity and expression in peripheral blood mononuclear cells. More importantly, we find that the recovery of GRK activity and expression on POD1 is highly variable among patients, with some patients showing little or no recovery. This finding is interesting because a poor recovery of GRKs after CPB may slow the desensitization of GPCRs on mononuclear cells that propagate inflammation, thus contributing to postoperative morbidity. Testing this possibility would require a larger prospective study.

The main purpose of our study was to determine if the acute increase in IL-6 seen after CPB is associated with a decrease in GRKs in peripheral blood mononuclear cells. We found that the CPB-induced increase in IL-6 had a negative correlation with GRK activity and GRK6 expression. However, our data suggest that while IL-6 may contribute to the observed decrease in GRKs, these kinases recover in many patients on POD1 despite increased concentrations of plasma IL-6. This observation differs from the persistent decrease in GRK2 demonstrated by Lombardi et al.  after exposure of mononuclear cells to IL-6 in vitro . 18It seems that under in vivo  conditions, IL-6 may be the mechanism for the initial decrease in GRKs, but its effects are attenuated with time, possibly due to the release of antiinflammatory molecules such as IL-10. 5It should also be kept in mind that the observed effect on GRKs could be mediated through other cytokines or biomolecules since CPB is known to be accompanied with a profound inflammatory response. 5,6,16,17Clearly, further studies are needed to delineate the molecular events responsible for the observed decline in GRKs after CPB.

In conclusion, we have demonstrated that CPB is associated with a significant decrease in GRK activity as well as in GRK2 and GRK6 protein expression in mononuclear cells. The significant variability among patients in the decline and recovery of GRKs after CPB should be further investigated as potential markers for the development of postoperative complications. Finally, it should be pointed out that the decline in GRKs seen in mononuclear cells after CPB may or may not occur in other tissues.

The authors thank Yuri Klyachkin (Technician) for technical assistance and Kimberly B. Allison (Secretary) for editorial assistance (both at Department of Anesthesiology, Duke University Medical Center, Durham, NC).