MGMT protein has a molecular weight of ~26 kDa and antibodies against this protein are readily available e.g., https://www.cellsignal.com/products/primary-antibodies/mgmt-antibody/2739
While deciding the best loading control for a Western Blot, you should take into consideration the following factors.
1. Make sure that the loading control is expressed ubiquitously at high levels in your samples and that its level isn't affected by experiment conditions. For instance, if you are studying a metabolic process, use of GAPDH as a loading control is not recommended due its role in glycolysis.
2. Ensure that the loading control and your protein of interest are expressed in the same location. For instance, Tubulin, β Actin and GAPDH are whole cell proteins, the locality of PCNA is cell nuclear, Lamin B1 is located in nuclear membrane, and so on and so forth.
3. Choosing a loading control that is close in size, but still large or small enough to be easily distinguished from your target protein is crucial. If the target protein and loading control are of the same size, visualization will be impaired and data interpretation compromised. A size difference of at least 5 kDa between your protein of interest and the loading control will help ensure to distinguish the bands.
Since I do not know your protein of interest (whether it is cytoplasmic, nuclear, cytoskeletal or mitochondrial), I would not be able to give you the best controls with a small kDa size. Nevertheless, you could always refer to Table 1 and Table 2 of the link provided below, which gives you the common loading control proteins with their cellular localization and molecular weight.