Antibody-dependent cell-mediated cytotoxicity (ADCC).  Yes non-neutralizing antibodies can reduce viral loads.  They have the ability to label the virus for destruction (as you said, opsonize) without the absolute ability to neutralize infection, and worked hand in hand with the cellular immunity to orchestrate and direct the overall immune response (e.g. Fc receptors).  My previous lab worked on HIV vaccine development, which is well known to defer neutralization by rapid mutation of immunodominant domains. For the most part, neutralizing antibodies were unable to neutralize anything but the same virus (animal model) used to make the vaccine with (homologous), but we were still able to get significant cross protection in our model.    We could not test ADCC due to lack of commercial reagents, but we did see a trend towards cellular immunity as the major player, and large amounts of antibodies were produced with each vaccination.  ADCC is not inferior to neutralization, but just a bit more difficult to fully analyze, and less pronounced in a research/vaccine setting.  ADCC is an necessary function of an immune response to most pathogens, and it would stand to reason that those epitopes should be included in a vaccine.  Just make sure you are aware of immunodominance and evasion strategies.  Try not to bias the response towards a possible ineffective (or devious) epitope.

the answer is:  it depends.  it depends on the in vitro test used for the neutralization assay.  often there is only one cell type whereas, in the body, there are many.  so an antibody that can't block viral binding, for example, might still be able to opsonize and increase the frequency with which the virus is picked up by macrophages.   i can give you a case with anthrax.  the neutralization test is done by putting macrophages or a macrophage cell line into a dish, adding the serum sample, incubating for a while and then adding the anthrax toxin.  if the macrophages don't die, it's considered a positive test.  we did some experiments immunizing adult and baby mice with anthrax, and found that many of the mice had no neutralizing antibodies but were nevertheless protected from challenge with the toxin.  this is kind of sad, since the CDC has decided that this invitro test is the gold standard for knowing whether a person is immune,  and it often takes many injections into poor soldiers to reach a positive neutralization titer.  yet it seems to me that the test is not accurate.  yes, if you have a high titer, you're protected, but you could also be protected with a much lower titer.

so........for you virus experiments, the best thing i can suggest is to try your antibodies in vivo and see if they protect.  although many in vitro neutralization tests might be accurate, many others will not be.  do the in vivo test to be sure.

Antibody-dependent cell-mediated cytotoxicity (ADCC).  Yes non-neutralizing antibodies can reduce viral loads.  They have the ability to label the virus for destruction (as you said, opsonize) without the absolute ability to neutralize infection, and worked hand in hand with the cellular immunity to orchestrate and direct the overall immune response (e.g. Fc receptors).  My previous lab worked on HIV vaccine development, which is well known to defer neutralization by rapid mutation of immunodominant domains. For the most part, neutralizing antibodies were unable to neutralize anything but the same virus (animal model) used to make the vaccine with (homologous), but we were still able to get significant cross protection in our model.    We could not test ADCC due to lack of commercial reagents, but we did see a trend towards cellular immunity as the major player, and large amounts of antibodies were produced with each vaccination.  ADCC is not inferior to neutralization, but just a bit more difficult to fully analyze, and less pronounced in a research/vaccine setting.  ADCC is an necessary function of an immune response to most pathogens, and it would stand to reason that those epitopes should be included in a vaccine.  Just make sure you are aware of immunodominance and evasion strategies.  Try not to bias the response towards a possible ineffective (or devious) epitope.

ADCC would work if you had an antibody against a component of the virus that shows up on the membrane of the infected cell.  but would not work (that i know of) if the antibody were against a molecule exposed on the virion, but not on the cell.  still.......in a mixed antibody response, there should be some antibodies that work via NK cells and ADCC.

Bound antibodies can also induce the complement system (IgG and IgM). I am not sure however how it works on virus. It can certainly lead to opsonization.

One issue is, would opsonisation be an advantage to the virus? As happens with non-neutralising antibodies in antibody-dependant enhancement in dengue virus. The virus is more than happy to enter the cells and infect them.

Like the AntiHBs antibody that is persistent for years in the serum after HBV vaccination I assume both ADCC and neutralisation targets can be used as vaccine targets. (correct me if wrong)

answering both Terje and Vasishta:  whether complement works on the virus itself, and whether opsonization will help the virus, will depend entirely on the individual virus that you're using.  viruses are amazingly different one to the other.  for example, flu is a membraned virus and can be killed by complement and would not gain from being opsonized and taken up by macrophages.  other viruses have other characteristics.  so you need to know the physiology of the virus your'e working with.

Yes non-neutralising antibodies can do so. They have the ability of opsonizing virus particles without the absolute ability to neutralize infection. Fc receptors can direct the overa immune response as well.