gunicorn + zope error handling

we are trying to setup gunicorn + plone. It works well so far, but fails to handle errors (404, 500 etc) and throws 'Internal Server Error' while it should return plone's error page.

Example case of requesting a page that doesn't exist is pasted below. The question is how should gunicorn be setup so that errors are handled themselves by zope/plone, as other requests?

Cheers and regards

==> var/log/gunicorn-stdout---supervisor-JhaTfg.log <==
 2012-05-24 01:41:16 [15137] [ERROR] Error handling request
 Traceback (most recent call last):
   File "/home/user/testing/eggs/gunicorn-0.14.3-py2.6.egg/gunicorn/workers/sync.py", line 100, in handle_request
     respiter = self.wsgi(environ, resp.start_response)
   File "/home/user/testing/eggs/repoze.retry-1.0-py2.6.egg/repoze/retry/__init__.py", line 90, in __call__
     app_iter = self.application(environ, replace_start_response)
   File "/home/user/testing/eggs/repoze.tm2-1.0b2-py2.6.egg/repoze/tm/__init__.py", line 24, in __call__
     result = self.application(environ, save_status_and_headers)
   File "/home/user/testing/eggs/repoze.vhm-0.14-py2.6.egg/repoze/vhm/middleware.py", line 106, in __call__
     return self.application(environ, start_response)
   File "/home/user/testing/eggs/Zope2-2.13.13-py2.6.egg/ZPublisher/WSGIPublisher.py", line 255, in publish_module
     response = _publish(request, 'Zope2')
   File "/home/user/testing/eggs/Zope2-2.13.13-py2.6.egg/ZPublisher/WSGIPublisher.py", line 185, in publish
     object = request.traverse(path, validated_hook=validated_hook)
   File "/home/user/testing/eggs/Zope2-2.13.13-py2.6.egg/ZPublisher/BaseRequest.py", line 518, in traverse
     return response.notFoundError(URL)
   File "/home/user/testing/eggs/Zope2-2.13.13-py2.6.egg/ZPublisher/HTTPResponse.py", line 718, in notFoundError
     "<p><b>Resource:</b> %s</p>" % escape(entry))
 NotFound:   <h2>Site Error</h2>
   <p>An error was encountered while publishing this resource.
   </p>
 

---