**4.2 Differential membrane fusion activities of VAMPs**

Do VAMPs have differential membrane fusion capacities? An ideal experimental system to answer this question will require equal expression of VAMP proteins and a quantitative membrane fusion assay. Using anti-Myc staining and flow cytometry as measurement (Figs. 3C and D), we express VAMP proteins at the same level at the cell surface. Furthermore, using the quantitative enzymatic cell fusion assay, we compare their membrane fusion activities.

When COS-7 cells are transfected with the flipped SNARE plasmids at the same concentration, cell surface expression of VAMPs 5 and 8 is more than 2 fold higher than VAMPs 1, 3, 4 and 7, and cell surface expression of syntaxin4 is 1.8 fold higher than syntaxin1 (Hasan et al., 2010) (data not shown). To express the v- and t-SNAREs at the same level, we titrate and optimize the concentration of each flipped SNARE plasmid used in transfection. Flipped SNARE plasmids are transfected at the following concentrations (per 10 cm2 growth area, *i.e.*, per well in 6-well plates): VAMP1, 0.2 g; VAMP3, 0.5 g; VAMP4, 0.5 g; VAMP5, 0.05 g; VAMP7, 1.0 g; VAMP8, 0.1 g; syntaxin1, 0.5 g; syntaxin4, 0.05 g. tTA, TRE-LacZ and flipped SNAP-25 are cotransfected at 1 g per 10 cm2 growth area. Under such conditions, VAMPs 1, 3, 4, 5, 7 and 8 are expressed at same level at the cell surface, while syntaxins 1 and 4 are expressed at the same level (Fig. 3D). Because the flipped VAMP2 protein does not contain a Myc tag (Hu et al., 2003), we are not able to compare its expression level with the other VAMPs.

After expressing VAMPs 1, 3, 4, 5, 7 and 8 at the same level, we compare their membrane fusion activities using the enzymatic cell fusion assay. With syntaxin1/SNAP-25, VAMPs 1, 3, and 8 have comparable and the highest fusion activities, whereas VAMPs 4 and 7 have 50% and 30% lower fusion activities, respectively (Fig. 6A). With syntaxin4/SNAP-25,

VAMPs 1, 2, 3, 7 and 8 are known to drive vesicle fusion with the plasma membrane. The above data provide additional evidence that vesicles that carry either one of these five VAMPs are capable of fusing with the plasma membrane. Since these VAMPs are functionally redundant, they may compensate each other in loss of function studies

To rule out the possibility that the baseline -galactosidase activity detected in the VAMP5 combinations (Fig. 5) is caused by residual membrane fusion activity of VAMP5, we perform the microscopic cell fusion assay, which analyzes individual cell fusion events and can detect rare fusion events. In multiple experiments, no cell fusion is observed in the microscopic assay when VAMP5 is combined with syntaxin1/SNAP-25 or syntaxin4/SNAP-25, whereas VAMP4 drives fusion efficiently with both t-SNARE complexes (Hasan et al., 2010) (data not shown). Together, the results using enzymatic and microscopic cell fusion assays suggest that VAMP5 is unable to mediate membrane fusion

VAMP5 is expressed in muscle cells, in which it is mainly associated with the plasma membrane as well as intracellular vesicles (Zeng et al., 1998; Zeng et al., 2003). It is still possible that VAMP5 forms fusogenic interactions with other plasma membrane t-SNAREs to mediate exocytosis in muscle cells. Indeed, SNAP-23, a ubiquitously expressed SNAP-25 homolog, is expressed in skeletal muscle (Bostrom et al., 2010). The role of VAMP5 in

Do VAMPs have differential membrane fusion capacities? An ideal experimental system to answer this question will require equal expression of VAMP proteins and a quantitative membrane fusion assay. Using anti-Myc staining and flow cytometry as measurement (Figs. 3C and D), we express VAMP proteins at the same level at the cell surface. Furthermore, using the quantitative enzymatic cell fusion assay, we compare their membrane fusion activities.

When COS-7 cells are transfected with the flipped SNARE plasmids at the same concentration, cell surface expression of VAMPs 5 and 8 is more than 2 fold higher than VAMPs 1, 3, 4 and 7, and cell surface expression of syntaxin4 is 1.8 fold higher than syntaxin1 (Hasan et al., 2010) (data not shown). To express the v- and t-SNAREs at the same level, we titrate and optimize the concentration of each flipped SNARE plasmid used in transfection. Flipped SNARE plasmids are transfected at the following concentrations (per 10 cm2 growth area, *i.e.*, per well in 6-well plates): VAMP1, 0.2 g; VAMP3, 0.5 g; VAMP4, 0.5 g; VAMP5, 0.05 g; VAMP7, 1.0 g; VAMP8, 0.1 g; syntaxin1, 0.5 g; syntaxin4, 0.05 g. tTA, TRE-LacZ and flipped SNAP-25 are cotransfected at 1 g per 10 cm2 growth area. Under such conditions, VAMPs 1, 3, 4, 5, 7 and 8 are expressed at same level at the cell surface, while syntaxins 1 and 4 are expressed at the same level (Fig. 3D). Because the flipped VAMP2 protein does not contain a Myc tag (Hu et al., 2003), we are not able to

After expressing VAMPs 1, 3, 4, 5, 7 and 8 at the same level, we compare their membrane fusion activities using the enzymatic cell fusion assay. With syntaxin1/SNAP-25, VAMPs 1, 3, and 8 have comparable and the highest fusion activities, whereas VAMPs 4 and 7 have 50% and 30% lower fusion activities, respectively (Fig. 6A). With syntaxin4/SNAP-25,

(Borisovska et al., 2005).

with the plasma membrane t-SNAREs.

exocytosis needs further investigation.

**4.2 Differential membrane fusion activities of VAMPs** 

compare its expression level with the other VAMPs.

VAMPs 1 and 3 have comparable and the highest fusion activities, whereas VAMPs 4, 7 and 8 have 36%, 26% and 54% lower fusion activities, respectively (Fig. 6B). As expected, only baseline β-galactosidase activity is detected when VAMP5 is paired with the t-SNAREs. These data indicate that VAMPs have differential membrane fusion activities with plasma membrane t-SNAREs. However, when expressed at higher levels, VAMP4 drives membrane fusion as efficiently as VAMPs 1 and 3 (Fig. 5). Since the differences of fusion activities among the fusogenic VAMPs are within a factor of 2, these results imply that with the exception of VAMP5, VAMPs are essentially redundant in mediating membrane fusion with plasma membrane t-SNAREs.

Fig. 6. Comparison of fusion activities of VAMPs. With optimized transfection conditions, VAMPs 1, 3, 4, 5, 7 and 8 are expressed at the same level at the cell surface, while syntaxin1/SNAP-25 (A) and syntaxin4/SNAP-25 (B) are expressed at the same level. Cell fusion is quantified using the enzymatic fusion assay. The fusion activities (OD420) of control cells (-VAMP) and the v-cells expressing VAMPs 3, 4, 5, 7 or 8 are normalized to the fusion activity of the v-cells expressing VAMP1. Error bars represent standard deviation of four independent experiments.

Analysis of SNARE-Mediated Exocytosis Using a Cell Fusion Assay 241

The log-log plot has been used to determine the cooperativity of viral fusion proteins in membrane fusion (Danieli et al., 1996). Using log-log plot, we further analyze the cooperativity of VAMP1 proteins. If two VAMP1 (V) proteins take part in the cell fusion reaction, *i.e.*, V + V → Fusion, the rate of fusion (F) = k [V]2. Therefore, log (F) = log (k) + 2 log [V], and the slope of the resulting log-log plot will be 2. A log-log analysis of the dependence of cell fusion activity on VAMP1 cell surface density is shown in Fig. 7B. Linear regression is performed to model the log-log correlation (Fig. 7B), and the resulting slope is 0.52. These results further suggest that the cell fusion reaction does not involve concerted

Previous studies have estimated that 1 to 11 SNARE complexes are needed for membrane fusion (Domanska et al., 2009; Hua and Scheller, 2001; Karatekin et al., 2010; van den Bogaart et al., 2010). In the cell fusion reaction, we do not observe cooperativity of VAMP1 proteins, suggesting that concerted action of multiple SNARE complexes is not required to fuse cellular membranes. However, to achieve fast exocytosis in intact cells, concerted action of multiple SNARE complexes is clearly needed (Mohrmann et al., 2010). Such cooperativity of SNARE complexes may be organized by the binding of regulatory proteins such as

To examine v-/t-SNARE interactions quantitatively, we developed an enzymatic cell fusion assay that utilizes activated expression of β-galactosidase and spectrometric measurement. Using this assay, we show that VAMPs 1, 2, 3, 4, 7 and 8 mediate membrane fusion efficiently with plasma membrane t-SNAREs syntaxin1/SNAP-25 and syntaxin4/SNAP-25, whereas VAMP5 does not drive fusion with the t-SNAREs. By expressing VAMPs 1, 3, 4, 7 and 8 at the same level, we further compare their membrane fusion activities. VAMPs 1 and 3 exhibit comparable and the highest fusion activities, whereas VAMPs 4, 7 and 8 have 30 - 50% lower fusion activities. Collectively, these data indicate that VAMPs have differential membrane fusion activities, and imply that with the exception of VAMP5, VAMPs are essentially redundant in mediating membrane fusion with plasma membrane t-SNAREs. Furthermore, no cooperativity of VAMP1 proteins is observed in the cell fusion reaction, suggesting that concerted action of multiple SNARE complexes is not required to fuse

We thank Adrienne Bushau and David Humphrey for critically reading the manuscript. This work was supported by startup funds from the University of Louisville and CA135123

Advani, R.J., Bae, H.R., Bock, J.B., Chao, D.S., Doung, Y.C., Prekeris, R., Yoo, J.S., and

Scheller, R.H. (1998). Seven novel mammalian SNARE proteins localize to distinct membrane compartments. *Journal of Biological Chemistry*, Vol.273, pp. 10317-10324.

synaptotagmins and Munc18 (Chicka et al., 2008; Shen et al., 2007).

action of VAMP1 proteins.

**5. Conclusion** 

cellular membranes.

**7. References** 

**6. Acknowledgement** 

from the National Institutes of Health (to C.H.).

#### **4.3 Cooperativity of VAMP proteins in the cell fusion reaction**

The number of SNARE complexes that cooperate to mediate vesicle fusion is under active investigation. To investigate the cooperativity of VAMPs in membrane fusion, we determine the dependence of cell fusion activity on cell surface expression level of VAMP1. We choose VAMP1 in this experiment because it has high membrane fusion activity (Fig. 6). In t-cells, the cell surface expression levels of syntaxin1 and SNAP-25 are kept constant. v-Cells are transfected with increasing concentrations of the flipped VAMP1 plasmid. At each concentration, we measure the cell surface expression level of VAMP1 proteins using flow cytometry, and determine cell fusion activity of VAMP1 with syntaxin1/SNAP-25 using the enzymatic fusion assay. Cell fusion activity is then plotted as a function of the mean fluorescence intensity of VAMP1 staining (Fig. 7A). The correlation is best fit with a polynomial regression. The hyperbolic instead of sigmoidal correlation (Fig. 7A) suggests that there is no cooperativity of VAMP1 proteins in driving cell fusion.

Fig. 7. Dependence of cell fusion activity on cell surface density of VAMP1. (A) v-cells expressing increasing amount of VAMP1 at the cell surface are combined with t-cells expressing syntaxin1/SNAP-25. Cell fusion activities are quantified and correlated with the mean fluorescence intensity of VAMP1 staining. (B) Log-log plot of cell fusion activity vs. mean fluorescence intensity of VAMP1 staining.

The log-log plot has been used to determine the cooperativity of viral fusion proteins in membrane fusion (Danieli et al., 1996). Using log-log plot, we further analyze the cooperativity of VAMP1 proteins. If two VAMP1 (V) proteins take part in the cell fusion reaction, *i.e.*, V + V → Fusion, the rate of fusion (F) = k [V]2. Therefore, log (F) = log (k) + 2 log [V], and the slope of the resulting log-log plot will be 2. A log-log analysis of the dependence of cell fusion activity on VAMP1 cell surface density is shown in Fig. 7B. Linear regression is performed to model the log-log correlation (Fig. 7B), and the resulting slope is 0.52. These results further suggest that the cell fusion reaction does not involve concerted action of VAMP1 proteins.

Previous studies have estimated that 1 to 11 SNARE complexes are needed for membrane fusion (Domanska et al., 2009; Hua and Scheller, 2001; Karatekin et al., 2010; van den Bogaart et al., 2010). In the cell fusion reaction, we do not observe cooperativity of VAMP1 proteins, suggesting that concerted action of multiple SNARE complexes is not required to fuse cellular membranes. However, to achieve fast exocytosis in intact cells, concerted action of multiple SNARE complexes is clearly needed (Mohrmann et al., 2010). Such cooperativity of SNARE complexes may be organized by the binding of regulatory proteins such as synaptotagmins and Munc18 (Chicka et al., 2008; Shen et al., 2007).
