You want your regular expression to match many times (g) and look past the 1st newline (m):

use strict; 
use warnings; 

my $data = '>protein_name1
blablabla
>protein_name2
blablabla
>protein_name3
blablabla
';

my (@header) = $data =~ /^>(.+)/gm; 
print "@header\n";

__END__

protein_name1 protein_name2 protein_name3
[download]

• perlre
• http://www.bioperl.org/wiki/Main_Page?