跳到内容
现在

科学 · 人类骨骼演化

人类软骨中的保水糖分更少。基因提供了线索

一项人类与类人猿的比较，将基因调控变化与软骨中保水GAG水平降低联系起来。但这种变化对骨关节炎有多大影响，仍未有定论。

文字 阅读需 6 分钟 · 19 个来源

HeyDex

原语言视频

译自英文原文。

人类骨骼的进化史通常围绕其成就来讲述：直立行走、长腿，以及容纳大脑发育的空间。较少受到赞誉的特征，则包括可能变得疼痛难用的关节。如今，一项将人类特有的遗传变化与软骨化学联系起来的研究，为这段漫长历史与一种常见残疾之间提供了一个合理的关联——但并未证明膝盖不好是成为人类所付出的代价。⁠3⁠13

核心发现引人注目。在8种关节中，人类软骨样本所含的硫酸化糖胺聚糖（GAG）显著少于其他类人猿样本。这些长链糖分子有助于软骨保水并承受压缩。按DNA含量校正后，人类的水平低2.6至4倍，整体差异被描述为大约三倍。这是组织化学上的差异，并非对骨关节炎发生可能性增加幅度的测量。⁠3

这项研究于2026年9月23日发表于《自然》，并因11月14日《耶路撒冷邮报》对共同负责人David Gokhman的采访而再次受到关注。该研究由他在以色列魏茨曼科学研究所的实验室与京都大学Fumitaka Inoue的实验室合作完成，Nadav Mishol和Yizhi Yan负责牵头研究工作。11月的报道带来的是一次采访，而非新发表的实验。⁠3⁠12⁠1

8人类GAG水平较低的关节类型 3
42研究人员报告的人类软骨捐献者人数 3
7组织比较中的类人猿个体数 3

GAG之所以重要，是因为软骨的缓冲作用不仅取决于软骨细胞。软骨细胞会生成细胞外基质——围绕并支撑它们的材料。GAG链带有很强的负电荷，有助于将水分吸引到水合网络中，从而增强弹性和抗压能力。GAG约占软骨干重的15–30%；这个比例不包括水分本身。⁠13

研究人员提出，生成减少会使人类的生化储备变小，而进一步流失可能由此促成退化。这是一个有用的假说，正因为它明确指出了值得研究的对象。但它无法解释每一个疼痛关节，不能证明衰退不可避免，也没有显示恢复到类似类人猿的水平就能预防疾病。⁠3

一致的模式，小规模的比较

组织研究涵盖髋、膝、肩、肘、腕、指关节和两类手指关节。研究人员使用染料结合检测法测量硫酸化GAG，随后按DNA含量对测量结果进行标准化。他们没有直接测量关节的总体缓冲能力、软骨体积或各关节的机械强度。⁠3

样本包括111份来自42名人类捐献者的组织，以及28份来自7只类人猿的组织。这些区别很重要：来自同一个体的多块软骨，不会因此变成多个独立的捐献者。在总体比较中，重复测量结果被取平均值，每个个体只计一个数值。有些指关节比较只有一只类人猿提供组织，这大大限制了这些特定结果代表更广泛种群的能力。⁠3

尽管如此，每一类关节都指向相同方向。对年龄和性别进行调整后，这种差异依然存在；使用两种方法在物种间匹配年龄后，差异也依然存在。差异还出现在人类承受不同机械负荷的关节中。这使得仅用某个特定关节的使用量来解释差异显得不那么有说服力，但并未排除所有环境影响。⁠3

所有标本均在死亡后采集；研究没有为此杀死任何动物。人类组织来自Technion的自愿捐献遗体项目，动物园方面的合作则提供了类人猿材料。采集条件并不完全相同：人类标本经过化学保存，而指定的黑猩猩手部材料则经过冷冻。这是一项小规模横断面比较，并非记录人类和类人猿软骨在一生中如何变化，也没有直接测量这种差异在人类发育过程中何时开始。⁠3

通向同一条通路的不同路径

遗传学研究为这种化学差异提供了一种可能的机制。研究团队没有从某个疑似关节炎基因入手，而是比较了15,708名现代人类和139只类人猿的基因组。3个尼安德特人基因组和1个丹尼索瓦人基因组帮助研究人员识别现代人类与古代人类谱系共同具有的变化。在这项搜索中，研究人员选出了561,410个候选调控区域中的单字母替换：这些DNA序列有助于控制基因何时、何地以及以多大程度发挥作用。⁠3

研究人员构建了这些区域的人类版本和祖先版本，并在人体软骨细胞中测试其活性。在显示活性的65,576对序列中，有15,077对（23%）在两个版本之间表现出显著差异。大多数人类版本和祖先版本的表现相似。而在这项检测中没有活性的序列，仍可能在其他组织或发育阶段发挥作用。⁠5

另一项实验则将人类和类人猿的染色体置于相同的实验室细胞中。研究人员将培养的人类细胞与黑猩猩或大猩猩细胞融合，然后引导这些杂交细胞朝着能够形成骨骼和软骨的早期阶段发育。这些是杂交细胞，不是杂交动物。共享的细胞环境有助于区分邻近调控DNA的影响与物种分别在不同培养物中培育时产生的差异。⁠3⁠12

两种方法都指向构建和维持细胞外基质的机制。在杂交细胞分析中，21个活性存在差异的GAG相关基因中，有17个人类一侧的活性较低。这一结果提示，变化协调地影响了整个生成通路，而非某个单一突变关闭了软骨生成。⁠3

研究还发现一个重要的复杂因素：人类ACAN基因发生了扩增。该基因编码聚集蛋白聚糖支架，可承载许多GAG链。人类版本平均多出6个重复，从而增加了12个潜在附着位点。研究人员提出，这些额外位点可能补偿GAG生成减少，但这一点尚未得到证明。更多附着链的位置并不一定意味着生成的链更多——这正是直接测量组织不可或缺的原因。⁠3

进化的权衡仍未得到证明

软骨GAG减少已与骨关节炎及其他退行性疾病相关。研究人员提出，物种层面的减少会增加人类的易感性，而在一生中GAG还会进一步流失。但这项研究没有追踪疾病发生率，也没有计算个人额外增加的风险。类人猿也会患关节疾病；这种比较并不是把人类的脆弱与类人猿的免疫力相对照。⁠3

这种变化为何得以延续，仍更难确定。一种假说涉及神经系统：GAG有助于稳定神经元之间的连接，因此较低的水平可能延长大脑可塑性并增加学习机会。另一种假说将生成减少与更轻、更纤细且适合耐力跑的骨骼联系起来。这项研究既没有测试学习能力，也没有测试跑步表现。这里没有证实任何一种优势。⁠13

并非每一种持续存在的进化变化都必定代表成功的权衡。论文同时探讨了适应和负向选择减弱这两种可能性——也就是说，限制生成水平降低的约束可能有所减弱。因此，过早宣称自然选择以健康关节换取智力或耐力并不妥当。⁠3

对患者而言，界限同样明确。骨关节炎治疗中已经使用涉及GAG的注射，但这不是一项治疗试验，也没有证实较早接受注射可以预防残疾。研究没有产生经过验证的新筛查检测、预防方案或基因疗法。在接受《耶路撒冷邮报》采访时，Gokhman提醒说：“但距离将其用于基因治疗还很远。”⁠1⁠3

这项研究为研究人员提供了更具体的生物学机制和有待探索的潜在靶点。改变这一机制能否安全地保护行动能力，仍是一个未解问题。这项发现不是治疗关节疼痛的良方，而是一个有更扎实依据、值得检验的解释。

来源15

在此处引用的来源中，Google News Kenya (KE:en)最先发布，比本报道播出早20小时。
  1. Google News Kenya (KE:en)
  2. HeyDay News

1篇早期文章作为背景资料参阅。13家媒体未注明日期：NaturePubMed Central (PMC)doi.orgGitHubkyoto-u.ac.jpMcGill Law Journal

按媒体查看所有来源

来自 15 家媒体的 19 个来源· 引用最多的优先

播出原貌
  1. Human cartilage holds roughly a third as much water-trapping sugar material as ape cartilage.
  2. Osteoarthritis affects about 607 million people worldwide, and these molecules help joints withstand pressure. Researchers have traced a possible genetic reason for the difference, but how much does it explain our vulnerability to painful joints?
  3. Across eight joint types, human samples contained 2.6 to four times less sulfated GAG, after adjustment for DNA content. GAG stands for glycosaminoglycan, a family of long sugar chains that helps cartilage retain water.
  4. The chains carry a strong negative electrical charge. That helps them attract water into the material surrounding cartilage cells. The resulting hydrated network contributes to elasticity and resistance to compression. GAGs account for 15 to 30 percent of cartilage's dry weight, second only to collagen.
  5. Researchers at the Weizmann Institute of Science in Israel and Kyoto University in Japan combined genetic experiments with donated cartilage. Their five-year project connected changes in human DNA regulation to lower activity in the machinery that makes these water-holding molecules.
  6. The study was published September twenty-third, twenty twenty-six. A November interview brought it renewed attention. The experiments began with a broad search through genomes, looking for changes shared by humans rather than starting with one suspected arthritis gene.
  7. David Gokhman and Fumitaka Inoue co-led the collaborating laboratories. Nadav Mishol led analyses and manuscript work, while Yizhi Yan performed the large-scale sequence experiments. Together, their teams tested how human and ancestral versions of regulatory DNA behave in cartilage cells.
  8. The genome comparison included 15,708 modern humans and 139 great apes. Those apes included bonobos, chimpanzees, gorillas and orangutans. Three Neanderthal genomes and one Denisovan genome helped identify changes shared across modern and ancient human lineages.
  9. From an initial catalogue of about 5.7 million single-letter differences, the team selected 561,410 substitutions in candidate regulatory regions. These regions act like controls, influencing when and where genes work, and how strongly they operate.
  10. The researchers built 366,402 sequence pairs, comparing human and ancestral versions. Of those, 65,576 pairs showed activity. Within that active group, 15,077 regulatory elements behaved differently between versions, or 23 percent of the active elements.
  11. Each sequence was 270 DNA letters long. Nearby substitutions could share one sequence, so the sequence count differs from the substitution count. Genetic barcodes let researchers measure the activity of many sequences together. They tested the library in human cartilage cells, repeating the experiment three times.
  12. Most human and ancestral versions behaved similarly. And this experiment tested their activity in cartilage cells. A sequence that appeared quiet there could still have a role in another tissue or at another stage of development.
  13. The team also fused cultured human cells with chimpanzee or gorilla cells. They guided these laboratory hybrid cells toward an early stage capable of forming bone and cartilage. Human and ape chromosomes could then operate inside the same cellular environment.
  14. That shared environment helped researchers separate the effects of nearby regulatory DNA from differences between separate cultures. It also captured changes beyond single-letter substitutions, including inserted or deleted DNA. Both approaches pointed toward the network that builds and maintains the material around cartilage cells.
  15. Seventeen of 21 GAG-related genes with different activity were less active on the human side. The average decreases were also larger than the increases. The pattern points toward coordinated changes across a production pathway, rather than one mutation shutting cartilage production off.
  16. The sequence experiment supported that pattern. Genes involved in GAG metabolism appeared among relevant regulatory targets at 2.2 times the expected rate. One particularly clear example involved a gene that helps assemble chondroitin sulfate, one of cartilage's important sugar chains.
  17. Seven of nine changed regulatory elements linked to that gene acted downward. One human sequence produced 34 percent less activity. In hybrid cells, the human gene's expression was 42 percent below the chimpanzee version and 34 percent below the gorilla version.
  18. Another result pulled in the opposite direction. The human ACAN gene, which makes the aggrecan scaffold, had expanded. That scaffold carries many GAG chains. Human versions averaged six extra repeats, creating twelve additional potential attachment sites for those chains.
  19. The scaffold comparison used 316 human DNA versions and 26 great-ape versions. More attachment sites might compensate for lower GAG production, the researchers suggest. But those opposing changes made direct measurements of cartilage essential to learn the overall result.
  20. Subscribe to this channel as researchers investigate what this cartilage difference means for joint disease.
  21. The tissue measurements used 111 samples from 42 human donors, compared with 28 samples from seven great apes. Some donors supplied several samples. For the overall comparison, the researchers averaged those repeated measurements down to one value per individual.
  22. Human levels were lower in the hip, knee, shoulder, elbow and wrist. They were also lower at the knuckles and in two categories of finger joints. The researchers used a dye-binding test to measure sulfated GAG and adjusted the measurements for DNA content.
  23. All eight joint categories pointed in the same direction. But some finger comparisons included tissue from only one ape, limiting how precisely those results represent the wider population. Repeated laboratory readings improved measurement precision without adding more individual donors.
  24. The difference persisted after adjustments for age and sex. Two methods of matching ages across species also left ape samples higher. The pattern crossed joints that experience different loading in humans, making a simple explanation based only on a joint's immediate workload less persuasive.
  25. Every tissue sample was collected after death, and no animals were killed for the study. Consenting human body donors supplied tissue through the Technion. Zoological collaborations supplied ape material, including a chimpanzee hand from a female that died naturally at sixty.
  26. The human specimens were chemically preserved, while that chimpanzee hand had been frozen. And the comparison captured tissue at one point after death. It couldn't follow how each individual's cartilage changed over a lifetime.
  27. Lower GAG levels are already associated with osteoarthritis and degenerating spinal discs. The researchers propose that humans' lower production could leave a smaller biochemical reserve, with further losses contributing to degeneration through life. Great apes can develop joint disease too.
  28. The paper discusses an evolutionary GAG reduction of about 62 percent. Earlier studies found reductions of 25 to 38 percent in diseased versus healthy human tissue. Those separate comparisons don't identify a universal failure point. This experiment didn't measure the extra disease risk.
  29. The genetic experiments and tissue chemistry now fit together as a plausible mechanism for vulnerability. Why the change persisted during evolution is less certain. Researchers suggest that it could have accompanied advantages elsewhere, or that evolutionary constraints against lower production simply weakened.
  30. GAGs also help stabilize connections in the nervous system. Lower levels might have prolonged the brain's ability to change and learn. Another proposal links lower production to a more slender skeleton compatible with endurance running. Both explanations remain hypotheses.
  31. The researchers didn't test learning or running performance. Their findings therefore leave the particular evolutionary advantage unresolved. For patients, the immediate use is to give researchers more specific biological mechanisms and potential treatment targets to investigate.
  32. Injections involving GAGs are already used in osteoarthritis care, but this experiment didn't test whether earlier injections prevent disability. The atlas covers about 40 percent of candidate human enhancers, regions that help control genes, and could guide further work in cartilage and other tissues.
  33. Safely changing that regulatory machinery would require further research. The study offers no new treatment patients can receive today. David Gokhman put the distance to one possible application plainly. He said, “But we are far from using it for gene therapy.”

译文更新于

近期播出