Development of new amphiphilic agents for membrane protein study: Effect of hydrophobic density of detergent micelles on protein stability

Abstract

막 단백질은 유기체에 의해 생성된 단백질의 상당부분을 차지하고 세포 사이에 물질 및 정보 전달하는 것과 같은 많은 중요한 기능을 수행한다. 때문에 막단백질 연구는 현재 매우 중요한 연구 주제이지만 막단백질의 구조분석을 위해 지질 이중층으로부터 단백질을 추출하게 되면 외부환경에서 쉽게 변성되는 특성이 있어 그 연구가 매우 도전적이다. 전통적인 양친매성 분자들이 막단백질 연구를 위해 지금까지 많이 사용되어 왔지만 이 분자들에 의해 형성된 마이셀 안에서 막단백질들은 쉽게 응집되고 변성되었으며, 따라서 우리는 더욱 효과적인 화합물 개발이 필요하였다. 이 논문에서는 세가지 종류의 양친매성 분자에 대한 단백질 안정성 평가 결과를 소개하고 있으며 전략적으로 소수성기의 밀도를 조절하여 막단백질에 대한 특성을 크게 향상시켰다.

첫 번째 화합물인 TNM은 세 개의 친수성 그룹과 세 개의 소수성 그룹을 두 개의 네오펜틸 그룹을 이용하여 도입하였고, 두 네오펜틸 그룹 사이에 링커를 도입한 TNM-Ls과 링커 없이 바로 두 그룹을 연결한 TNM-Ss로 나뉜다. 총 10개의 TNM분자들을 합성하여 네 종류의 막단백질에 대해 평가하였고, 전반적으로 전통적인 양친매성 분자인 DDM보다 막단백질 안정화 측면에서 좋은 성능을 보여주었으며 각 단백질 안정화에 최적인 양친매성 분자는 단백질에 따라 달라지는 경향성을 보여주었다. 두 개의 새로운 분자들이 (TNM-C12L과 TNM-C11S) 다양한 막단백질에 대해서 DDM보다 우수한 특성을 보여주었다.

두 번째 화합물은 기존에 개발된 양친매성 분자인 LMNG의 소수성 그룹을 비대칭으로 디자인한 화합물이다. 즉 LMNG의 두 개의 친수성 그룹은 그대로 유지한 상태에서 두 개의 다른 길이의 알킬체인을 도입하여 비대칭성을 갖는 소수성기를 설계 및 제조하였다. 총 6개의 비대칭성 MNG (A-MNG)를 다양한 막단백질에 대해 테스트한 결과 기존의 LMNG보다 안정한 마이셀을 형성함을 알았고 막단백질 안정성에서도 좋은 효력이 있음을 알아내었다. 비대칭성 MNG중 MNG-6,14은 모든 막단백질에서 탁월한 효과를 보여주었으며 MNG-8,12는 GPCR 안정성에 우수하였다.

세 번째 화합물은 기존에 개발된 양친매성 분자인 OGNG를 구조적으로 변형하여 설계 및 제조하였다. OGNG는 작은 단백질-양친매성분자 결합체(PDC)를 형성하는 장점을 갖고 있지만 막단백질 안정화에 성능이 우수하지 못했다. 우리는 이를 개선하기 위해 기존의 OGNG의 친수성 그룹을 그대로 유지하면서 긴 알킬체인과 짧은 펜던트 체인을 소수성기로 도입하여 펜던트 사슬을 갖는 화합물들(P-GNGs)을 합성하였다. 많은 수의 P-GNG들은 기존의 OGNG 뿐만 아니라 DDM보다 더 안정한 마이셀을 형성하였고, 세 가지 P-GNG인 (GNG-2,14, GNG-3,13 그리고 GNG-3,14)은 모든 막단백질에서 DDM보다 우수한 성능을 보여주었다. 이와 같은 우수한 결과는 펜던트 체인의 존재로 인해 OGNG 마이셀과 비교하여 마이셀 내부의 소수성 밀도가 크게 증가한데 기인하는 것으로 보인다. 이를 통해 종합적으로 세 가지 종류의 양친매성 분자들 (TNMs, A-MNGs, P-GNGs)은 알킬사슬의 개수의 증가 또는 비대칭의 소수성기의 도입을 통해 소수성 밀도가 높은 마이셀을 형성하였고 이는 막단백질 안정화에 크게 기여할 수 있다는 것을 알 수 있다.

|Integral membrane proteins represent a large fraction of the proteins produced by living organisms and perform many crucial cellular functions, such as material transport and information exchange between cells and their environments. Therefore, membrane protein research is very important topic to advance understanding our biological system at the molecular level. However, it is a very challenging to study membrane proteins as these bio-macromolecules tend to easily denature or aggregate when extracted from the lipid bilayers for structural analysis of membrane proteins. Many amphiphilic agents have been developed to extract membrane proteins and stabilize them in aqueous solution. In micelles formed by many conventional detergents, membrane proteins undergo denaturation/aggregation and thus we have focused on development of amphiphiles effective at stabilizing membrane proteins. This thesis introduces three classes of amphiphilic agents that have detergent hydrophobic groups different from other detergents. Specifically, the new detergent class has multiple alkyl chains or asymmetric hydrophobic groups that increase detergent-protein and/or detergent-detergent interactions, leading to enhanced membrane protein stability.

The first class of amphiphiles, designated tandem neopentyl glycol maltosides (TNMs), contains three hydrophilic and three hydrophobic groups. One set has a propylene spacer between two neopentyl glycol units (TNM-Ls), while these two units were directly connected to each other in the other set (TNM-Ss). 10 TNMs with different alkyl chain length were prepared and evaluated with four membrane proteins. These new agents generally performed better than DDM at stabilizing the membrane proteins tested here. The best TNM tend to strongly depend on the target protein, indicating that a single agent is unlikely a magic bullet for all or even most membrane proteins.

The second class of amphiphiles contains the same head group as LMNG (i.e., a branched dimaltoside), but have two different alkyl chains. The difference in alkyl chain length was indicated in the detergent designation of asymmetric maltose neopentyl glycols (A-MNGs). In the variants, the number of total carbon units in the hydrophobic group was maintained as C20, which was attained by increasing one alkyl chain while decreasing the other alkyl chain by the same chain length. This class is the first example of introducing detergent asymmetric nature into novel detergent scaffold for membrane protein study. When six A-MNGs were tested with multiple membrane proteins, MNG-6,14 and MNG-8,12 showed an excellent effect on all membrane proteins tested here. In addition, MNG-8,12 was particularly effective at stabilizing two G protein-coupled receptors (GPCRs). The better packing of the detergent alkyl chains into the micelle interiors, originating from asymmetric detergent hydrophobic groups, is likely responsible for enhanced behaviors of these detergents with multiple membrane proteins.

The third class of amphiphiles was derived from OGNG, a recently developed novel detergent. This glucoside detergent tends to form small protein-detergent complexes (PDCs), a favorable feature for protein crystallization, but was limitedly effective at stabilizing membrane proteins. To overcome this issue, we designed and prepared pendant-bearing glucose neopentyl glycols (P-GNGs) with a long alkyl chain (main chain) and a short chain (pendant chain). The P-GNGs showed more stable micelle formation than OGNG and DDM. Of the P-GNGs, three P-GNGs (GNG-2,14, GNG-3,13 and GNG-3,14) were most superior to DDM for stabilizing all the tested membrane proteins. Consequently, the presence of the pendant chain, particularly the propyl chain, has a beneficial effect on protein stability, which is likely due to an increase in alkyl chain density in the micelle interior compared to OGNG or DDM micelles. High alkyl chain density achieved by introduction of the pendant chain would strengthen hydrophobic interactions between the detergent alkyl chains, resulting in enhanced micellar stability and membrane protein stability.

Combined together, three classes of novel detergents (TNMs, A-MNGs, and P-GNGs) were designed and synthesized for membrane protein study. These novel detergents contain the increased number of the alkyl chains or asymmetric hydrophobic groups, yielding high density of the detergent alkyl chains in the detergent micelle interiors. The novel detergents with such structural features conferred notably enhanced stability to multiple membrane proteins including GPCRs, thus representing high potential for membrane protein structural study. In addition, the detergent design principles discussed in these studies inspires many chemists to design novel detergents for membrane protein study.; Integral membrane proteins represent a large fraction of the proteins produced by living organisms and perform many crucial cellular functions, such as material transport and information exchange between cells and their environments. Therefore, membrane protein research is very important topic to advance understanding our biological system at the molecular level. However, it is a very challenging to study membrane proteins as these bio-macromolecules tend to easily denature or aggregate when extracted from the lipid bilayers for structural analysis of membrane proteins. Many amphiphilic agents have been developed to extract membrane proteins and stabilize them in aqueous solution. In micelles formed by many conventional detergents, membrane proteins undergo denaturation/aggregation and thus we have focused on development of amphiphiles effective at stabilizing membrane proteins. This thesis introduces three classes of amphiphilic agents that have detergent hydrophobic groups different from other detergents. Specifically, the new detergent class has multiple alkyl chains or asymmetric hydrophobic groups that increase detergent-protein and/or detergent-detergent interactions, leading to enhanced membrane protein stability.

The first class of amphiphiles, designated tandem neopentyl glycol maltosides (TNMs), contains three hydrophilic and three hydrophobic groups. One set has a propylene spacer between two neopentyl glycol units (TNM-Ls), while these two units were directly connected to each other in the other set (TNM-Ss). 10 TNMs with different alkyl chain length were prepared and evaluated with four membrane proteins. These new agents generally performed better than DDM at stabilizing the membrane proteins tested here. The best TNM tend to strongly depend on the target protein, indicating that a single agent is unlikely a magic bullet for all or even most membrane proteins.

The second class of amphiphiles contains the same head group as LMNG (i.e., a branched dimaltoside), but have two different alkyl chains. The difference in alkyl chain length was indicated in the detergent designation of asymmetric maltose neopentyl glycols (A-MNGs). In the variants, the number of total carbon units in the hydrophobic group was maintained as C20, which was attained by increasing one alkyl chain while decreasing the other alkyl chain by the same chain length. This class is the first example of introducing detergent asymmetric nature into novel detergent scaffold for membrane protein study. When six A-MNGs were tested with multiple membrane proteins, MNG-6,14 and MNG-8,12 showed an excellent effect on all membrane proteins tested here. In addition, MNG-8,12 was particularly effective at stabilizing two G protein-coupled receptors (GPCRs). The better packing of the detergent alkyl chains into the micelle interiors, originating from asymmetric detergent hydrophobic groups, is likely responsible for enhanced behaviors of these detergents with multiple membrane proteins.

The third class of amphiphiles was derived from OGNG, a recently developed novel detergent. This glucoside detergent tends to form small protein-detergent complexes (PDCs), a favorable feature for protein crystallization, but was limitedly effective at stabilizing membrane proteins. To overcome this issue, we designed and prepared pendant-bearing glucose neopentyl glycols (P-GNGs) with a long alkyl chain (main chain) and a short chain (pendant chain). The P-GNGs showed more stable micelle formation than OGNG and DDM. Of the P-GNGs, three P-GNGs (GNG-2,14, GNG-3,13 and GNG-3,14) were most superior to DDM for stabilizing all the tested membrane proteins. Consequently, the presence of the pendant chain, particularly the propyl chain, has a beneficial effect on protein stability, which is likely due to an increase in alkyl chain density in the micelle interior compared to OGNG or DDM micelles. High alkyl chain density achieved by introduction of the pendant chain would strengthen hydrophobic interactions between the detergent alkyl chains, resulting in enhanced micellar stability and membrane protein stability.

Combined together, three classes of novel detergents (TNMs, A-MNGs, and P-GNGs) were designed and synthesized for membrane protein study. These novel detergents contain the increased number of the alkyl chains or asymmetric hydrophobic groups, yielding high density of the detergent alkyl chains in the detergent micelle interiors. The novel detergents with such structural features conferred notably enhanced stability to multiple membrane proteins including GPCRs, thus representing high potential for membrane protein structural study. In addition, the detergent design principles discussed in these studies inspires many chemists to design novel detergents for membrane protein study.