First I used Perl::Tidy to reformat your script, making it a bit easier to ready. Then I began adding print statements so I could see what it was doing and fixing the errors.

Here is the modified script, which downloads and saves two postscript files which, I hope, is what you want it to be doing.

use strict;
use warnings;
use LWP::Simple;

my $url3 =
"http://www.ebi.ac.uk/thornton-srv/databases/cgi-bin/pdbsum/GetPage.pl
+?pdbcode=2j6p&template=ligands.html&l=1.1";
my $content = get($url3) or die "$url3: get failed";

use HTML::TreeBuilder;
my $p = HTML::TreeBuilder->new;
$p->parse_content($content);
my @href;
my $href;
my @anchors = $p->look_down( _tag => q{a} );
for my $anchor (@anchors) {
    my $txt = $anchor->as_text;
    if ( $txt =~ /EPE\s/ ) {
        print $txt, qq{\n};
        $href = $anchor->attr(q{href});
        print $href, qq{\n};
        chomp($href);
        push @href, $href;

#now I need to go to the link where there are my EPE ligand and then p
+arse and extract the link of the RunLigplot.pl that is the output of 
+a program LigPlot, written in perl, is a postcript, and I need that f
+ile as a script to extract info....I need to repeat these parsing for
+ every link,for every EPE....
    }
}
$p->delete;

for my $param (@href) {
    print "processing param: \"$param\"\n";
    my $url = "http://www.ebi.ac.uk$param";
    my $content = get($url) or die "$url: get failed";
    process_content($content);    # << This is missing from your code 
+!!!
}

sub process_content {
    my $content = shift;
    $p = HTML::TreeBuilder->new;
    $p->parse_content($content);
    my @href0;
    my @anchors0 = $p->look_down( _tag => q{a} );
    for my $anchor0 (@anchors0) {
        print "anchor: " . $anchor0->as_HTML() . "\n";
        my $href0 = $anchor0->attr(q{href});
        next unless($href0);
        print "found href: $href0\n";

        if ( $href0 =~ /(ligplot\d\d_\d\d)/ ) {
            my $id = $1;
            print $href0, qq{\n};
            push @href0, $href0;
            for my $param0 (@href0) {
                my $url = "http://www.ebi.ac.uk$param0";
                print "getting $url\n";
                $content = get($url) or die "$url: get failed";
#                my $content = shift;
#                print my $param0;

                #I need to download files from every link $param0...
                my $pdb = 'testpdb';
                my @files =
                  ( [ "http://www.ebi.ac.uk$param0", "$pdb.$id.ps" ], 
+);
                for my $duplet (@files) {
                    print "mirror $duplet->[0] to $duplet->[1]\n";
                    mirror( $duplet->[0], $duplet->[1] );
                }
            }
            $p->delete;

        }
    }

}
[download]

And here are the changes, other than the reformatting done by Perl::Tidy.

< my $content = get($url3);
---
> my $content = get($url3) or die "$url3: get failed";
32c32,34
<     my $content = get("http://www.ebi.ac.uk$param");
---
>     print "processing param: \"$param\"\n";
>     my $url = "http://www.ebi.ac.uk$param";
>     my $content = get($url) or die "$url: get failed";
42a45
>         print "anchor: " . $anchor0->as_HTML() . "\n";
43a47,48
>         next unless($href0);
>         print "found href: $href0\n";
45c50,51
<         if ( $href0 =~ /ligplot\d\d_\d\d'/ ) {
---
>         if ( $href0 =~ /(ligplot\d\d_\d\d)/ ) {
>             my $id = $1;
49,51c55,59
<                 $content = get("http://www.ebi.ac.uk$param0");
<                 my $content = shift;
<                 print my $param0;
---
>                 my $url = "http://www.ebi.ac.uk$param0";
>                 print "getting $url\n";
>                 $content = get($url) or die "$url: get failed";
> #                my $content = shift;
> #                print my $param0;
53a62
>                 my $pdb = 'testpdb';
55c64
<                   ( [ "http://www.ebi.ac.uk$param0", "$pdb.$param0.p
+l" ], );
---
>                   ( [ "http://www.ebi.ac.uk$param0", "$pdb.$id.ps" ]
+, );
56a66
>                     print "mirror $duplet->[0] to $duplet->[1]\n";
[download]