To clarify the question. Are you talking about serum samples after vaccination?  It does depend on how long you wait after immunisation.  I wonder whether it would be useful first to see if you have antibody by ELISA and then go on to test the function of the antibody using HAI.  We routinely use micro neutralisation assays since we got positives with that assay and we didn't need receptor destroying enzyme. 

Further clarification needed: what virus preparation and what kind of RBC are you using in the HAI? I assume you have included a positive control.

That depend. I think if you have just HA proteins (not a multimer of HA) they adsorb to RBC surface but they do not form network among them; therefore you would not see the hemagglutination effect. If you have a multimer subunit ( a subunit with more HAs) and they are functional, you should be able to generate a network among different RBCs and therefore to see the hemagglutination effect. Let's see if true ;)

Suggest you have a positive control to rule out suboptimal assay condition.  If  red cells you use (Turkey?)  is known to agglutinate  the starin of HA (sub unit vaccine here is HA?) you used is something to think about.  Buy commercial HA for comparison?

yes you can  have   an HA  with the subunit

if your initial viral  preparation titers  high  and if you obtain  sub unit with tween or othe r detergent ,evidently  in the first dilution there is a hemolysis   but  if you dilute more

 you can  view the HA you can absorb the detergent  before

no

I have performed many HA assay for the split vaccine, but not the subunit.

Detergent treated virus might give an HA after removal of the detergent, but depends on virus subtype and preparation. So, the answer to your question is: it depends.

Jonges M, Liu WM, van der Vries E, Jacobi R, Pronk I, Boog C, Koopmans M,

Meijer A, Soethout E. Influenza virus inactivation for studies of antigenicity

and phenotypic neuraminidase inhibitor resistance profiling. J Clin Microbiol.

2010 Mar;48(3):928-40.

http://jcm.asm.org/content/48/3/928.long

Haeagglutination reactions take place due to lattice formation (cross linking) between multiple antigen sites (influenza haemaggluinin, HA, in this case) on multiple influenza virus particles and receptors on multiple RBSc. HA of influenza viruses occurs in rosette form (multiple of trimeric HA). When numerous HA rosettes are present on the surface of virus, these bind to HA receptors on multiple RBCs, which in turn bind to more virus particles (as there are numerous receptors on RBCs), thus forming rosettes and visible agglutinates. In an influenza split virus preparation, HA is present as trimer, some rosettes, but not enough to form lattice or cross-linking of RBCs. Influenza split virus preparations are heterogenous in terms of amount of rosettes or trimeric HA and residual detergent. Therefore some preparations may show haemagglutination, others may not. Removing detergent may cause rosette formation and may result in haemaggluitnation. It seems that the preparation you have does not have enough rosettes to show agglutination. However, you could titrate the HA activity of your preparation by blocking the haemagglutination inhibition (HAI) activity of antibodies. In this case you take an antibody preparation with a known HAI titer and mix your subunit preparation with that. Now when you evaluate HAI titer of the antibody mixed with your preparation, the HAI titer will decrease. Of course, you have to take the virus preparation for the HAI test from the same or similar strain as your subunit HA preparation.

Thanks a lot everyone for your interesting suggestions/answers. I have A/california/07/2009 inactivated influenza subunit vaccine. It has been inactivated by formalin and cetyltrimethylammoniumbromide (CTAB) has been used as detergent. The final prept'n consists of

Formalin inactivation will result in a decrease in HA titre. this loss of HA titre can vary between preparations and could reach log10 value.

Formalin inactivation and effect on antigenicity has been described in our paper: Jonges M, Liu WM, van der Vries E, Jacobi R, Pronk I, Boog C, Koopmans M, Meijer A, Soethout E. Influenza virus inactivation for studies of antigenicity and phenotypic neuraminidase inhibitor resistance profiling. J Clin Microbiol. 2010 Mar;48(3):928-40. As well as inactivation by BPL, detergent (Triton X-100) and heat of whole virus. HA acitivity was well preserved after formalin inactivation of whole virus. So, the critical question is whether you have enough rosettes. Maybe put the preparation under the electronmicroscope to have a visual of what you have.

Haemaggltuination is one of the methods to evaluate biological activity of HA preparations. This is not strain specific and is influenced by a number of factors, such as unique characteristics of strain, which can affect the species from which RBCs can be agglutinated - turkey, guinea pig, chicken, horse, human 'O' type, etc. The antigenicity or potency method for HA preparations accepted by regulatory agencies is single radial immunodiffusion (SRID). You can get standard antigen and antibodies for this test from CBER, FDA or NIBSC, UK. HA from H1N1 strain, A/California/07/2009 has some unique characteristics, which can affect its SRID potency and/or HA activity depending upon the way it was purified and inactivated.

It seems logical you can't get positive hemmaglutination test with subunit vaccine because you have splitted virus with hemmaglutinins that are not able to form network with red blood cells. To get hemmaglutination you need to have whole virus which have many hemmaglutinin molecules on its surface and can bind more red blood cells ar the same time. However if you want to determine the potency of influenza vaccine this method is not accepted, single radial immunodiffusion is more reliable and recomended as official method for potency testring in Europen and other Pharmacopoeias.