One way I can think of is to dilute the mixture in a low ionic strength buffer to get the viscosity down, then apply it to an ion-exchange chromatography column. HPMC, being uncharged, will not bind to the column and will flow through. The protein will bind to the column, hopefully, and can be eluted with a high-salt wash.

Thanks Adam, it's a very good idea. I am trying to quantitate a protein in the matrix containing HPMC, so diluting the sample may decrease the quantitation limit. Also, if possible, a comparatively simple sample preparation method will be appreciated.

Is there huge difference in MW between your HPMC and your protein. MW cut off filters also another option after diluting it. 

The MW of HPMC E15 is about 50kDa, and the protein is about 17kDa, the filters are possible options. But what observed for PES membrane, which was used for filtration, is that HPMC is retained in the membrane, while the protein is also retained, in a low concentration range. 

Dilution of the protein sample should not be a problem in the method I suggested because the protein can be eluted in a concentrated form with a high-salt wash. To simplify sample processing, you could use pre-made ion-exchange resin-filled 96-well plates such as these (http://www.gelifesciences.com/webapp/wcs/stores/servlet/CategoryDisplay?categoryId=1106675&catalogId=66601&top=Y&storeId=11787&langId=-1), or make something similar yourself with bulk resin.