formula issue in bivariate logistic

[econcreek]

For the model blogit, when i have different covariates for the two dependent variables, the code returns "Error in formula.default(object, env = baseenv()) : invalid formula". The simulate code is as follows
n=100;
x1=rnorm(n);
x2=rnorm(n);
x3=rnorm(n);
y1=rbinom(n,1,0.5);
y2=rbinom(n,1,0.5);
mydata=data.frame(y1,y2,x1,x2,x3);
breg=zelig(list(y1x1+x2,y2x1+x3),model="blogit",data=mydata).

Anyone has the same issue of estimate bivariate logit model with different regressors?