Firstly, I would confirm gene expression in your samples by qPCR. 

To help more effectively, a picture of the blots you've generated so far would be helpful. 

Here are some questions that I ask myself when I get a bad blot with lots of NS binding:

Did I block correctly for protein of interest? Sometimes 5% skim milk in TBST is a bad choice. It tends to produce a lot of background with phospho-specific antibodies, for example. Try 5% BSA in TBST or donkey serum in TBST. 

Did I use too much primary, or incubate for too long? Try cutting the amount of primary in half, and try incubating in primary for 1 hr at 4'C with shaking instead of O/N. 

Did I use enough tween in my wash buffer? Sometime increasing the tween concentration can discourage non-specific binding. Try TBS with 1% Tween-20. 

Did I rinse effectively? Try washing 3x for 5 minutes each, followed by one 15 minute wash. This tends to produce better blots vs abbreviated wash steps. 

Am I using way too much secondary, or is the secondary binding non-specifically? Try titrating the secondary by dot-blots of primary antibody on unblocked membrane for the former. For the latter, try staining a blot with secondary only to see if you get the same pattern of non-specific binding. 

Is the primary antibody simply bad? Optimize your protocol using appropriate positive controls, and send your data suggesting a bad antibody to the manufacturer. They'll usually fix it at no expense to you. 

Firstly, I would confirm gene expression in your samples by qPCR. 

To help more effectively, a picture of the blots you've generated so far would be helpful. 

Here are some questions that I ask myself when I get a bad blot with lots of NS binding:

Did I block correctly for protein of interest? Sometimes 5% skim milk in TBST is a bad choice. It tends to produce a lot of background with phospho-specific antibodies, for example. Try 5% BSA in TBST or donkey serum in TBST. 

Did I use too much primary, or incubate for too long? Try cutting the amount of primary in half, and try incubating in primary for 1 hr at 4'C with shaking instead of O/N. 

Did I use enough tween in my wash buffer? Sometime increasing the tween concentration can discourage non-specific binding. Try TBS with 1% Tween-20. 

Did I rinse effectively? Try washing 3x for 5 minutes each, followed by one 15 minute wash. This tends to produce better blots vs abbreviated wash steps. 

Am I using way too much secondary, or is the secondary binding non-specifically? Try titrating the secondary by dot-blots of primary antibody on unblocked membrane for the former. For the latter, try staining a blot with secondary only to see if you get the same pattern of non-specific binding. 

Is the primary antibody simply bad? Optimize your protocol using appropriate positive controls, and send your data suggesting a bad antibody to the manufacturer. They'll usually fix it at no expense to you. 

Carl Manner pretty much have your Western Blot 101 covered :).

My experiences with detecting protein from tissues is not always easy, as you are trying to detect low abundant of protein of interest and the back ground usually masked your bank of interest or that you see the band of interest but you have to adjust the exposure time to high to detect your band of interest which also resulted in high noise-signal ratio.

Once you confirm that your antibodies are good, and you sure that your protein of interest is expressed in your tissue of interest.  I would recommend you do a pull down (capture) of your protein of interest from the total cell lysate, thereby increase your protein of interest concentration and reduce non-specific binding of your antibodies.

what system of signal detection you use? As there could be room for optimisation as well.  for example, you would never use Milk or PVDF with Odyssey system.

If it is a membrane protein that is not very abundant, it might be useful to do a plasma membrane extraction and then western blot. 

Prepare membrane without detergent and dissolve membrane pellet in 0.2-2% dodecyl maltoside, then run western blot

 Thank you everyone for your suggestions.

I am attaching my blot image for reference. i have marked a 70 kDa band in it. But above and below it also i have other bands. how can i get rid of them? 

Although I haven't worked with this particular protein, we do use Westerns for G-protein coupled receptors such as GLP-1R, which is also glycosylated and run as multimer (due to variable glycosylation). Since your protein is also glycosylated, I guess it may also run as multimer in Western. Please do the following to differentiate non-specific bands from the specific ones.

1. Use a positive control (the lysate of cells over-expressing this protein) to confirm you can detect your protein of interest in the immunoblotting.

2. If you get a single band in the positive control but several bands (as shown in the above blot) with your cells, try to partially purify the transporter by immunoprecipitation or cell surface biotinylation coupled with streptravidin bead pulling. The partial purification may get rid of some or all of the non-specific bands.

I have been working with taurine transporter for over a year now.

Here are my few suggestions:

Use 5% BSA in PBST for blocking, and I typically block for 2-3 hours.

do not heat the lysates before loading. I use 37'C for 20 min. The higher molecular wt that you see could be aggregates of the protein of interest as a result of boiling the samples. That is they form dimers which would be typically 140 KDa.

I am using antibody from thermofisher, which just need dilution of 1:4000 to detect taurine transporter.

Thank You for suggestion Madhur Shetty.