The human malaria parasite Plasmodium falciparum interacts with numerous polymorphic molecules in its host red blood cell (RBC), which presents difficulties in examining the effect of single RBC determinants on parasite infection. A good illustration is the story of fetal hemoglobin (HbF). First explored by two of the authors nearly five decades ago, the question of whether HbF inhibits invasion and/or growth continues to arise due to acquisition of epidemiologic data suggesting an interaction between HbF and sickle hemoglobin (HbS) that may influence malarial infection.^{1, 2} Here, we detail the studies investigating HbF's effect on malarial infection, highlighting the complexities associated with interpretation and providing suggestions of how to investigate further polymorphisms in parasitized RBCs from natural populations.

In foundational studies using experimental culture in vitro of infected mixed umbilical cord and HbAA cells, Pasvol et al. demonstrated that P. falciparum has a predilection for younger RBCs.³ This was done by analyzing blood smears from patients and cultures treated with a modification of the acid-elution-differential staining method of Kleihauer et al., a procedure that distinguishes parasitized HbF cells (F-cells) from normal (HbAA) cells since HbF is resistant to elution at pH 3.5.⁴ Thus, cells with differing hemoglobins could be tested under identical conditions rather than in parallel cultures. A higher percentage of F-cells, the younger and more metabolically active cells in the cord and HbAA blood mixture, were invaded by parasites. In addition, younger stages of parasites were seen in F-cells in the single cord blood sample used. Compared to the complete stalling of growth in sickle cell trait RBCs,^{5, 6} the growth delay was slight but significant. Lastly, Pasvol et al.³ confirmed that P. falciparum preferentially invades younger RBCs and not F-cells in infant blood samples. The younger cells in infant blood, that is, those that are HbA dominant, were more often invaded in up to two cell cycles. While these studies clearly demonstrated an effect of age and not HbF on parasite invasion, a slight but measurable effect of HbF on growth, was noted.

A subsequent study using mixtures of HbF and HbA dominant RBCs again confirmed that invasion was not influenced by HbF but by age of the RBCs.⁷ In addition to the cord-adult blood mix and blood of infants aged 3–6 months used in the above experiments, invasion into blood of heterozygotes of African Hereditary Persistence of Fetal Hemoglobin (HPFH) and a mix of blood of homozygotes for British HPFH and adult HbAA blood were studied. Once again, the younger cells (cord and HbAA RBCs) in the first two samples, respectively, were preferentially invaded. In the two HPFH samples, in which all RBCs were of the same age, there was no preference for either HbF or HbA dominant RBCs. Pasvol, Weatherall and Wilson also published another set of studies highlighting the preferential invasion of reticulocytes by P. falciparum.⁸ In these experiments, centrifugation was followed by isolation of RBCs from the top, middle and bottom of Wintrobe tubes. RBCs were then stained for reticulocytes with brilliant cresyl blue and counterstained for parasites with Giemsa's stain. Compared to normocytes, a higher parasitemia was found in reticulocytes and the top fraction that contains them. Wilson et al. also showed that the size of parasites in RBCs from a single culture of cord and adult blood were smaller in the F-cells.⁹ These studies again showed that HbF had no influence on invasion but suggested an influence on growth.

In these early studies, the focus was primarily the preferential parasitic invasion of younger cells with brief comments regarding delayed parasite development in F-cells. As a result, other studies^{10, 11} were conducted particularly focused on a mechanism for the HbF induced development delay. Friedman et al. could only show inhibition of parasite growth in HbF dominant RBCs (unspecified RBC type) in vitro in glutathione reduced media (medium 199), the same medium used in the Pasvol and Wilson studies. In RPMI (Roswell Park Memorial Institute) media, normal growth was observed. This suggested a mechanism of oxidant stress.¹⁰ Later, Shear et al., demonstrated a protective effect of HbF in vivo in a transgenic rodent malaria model with humanized alpha and gamma globin chains.¹¹ When infected with P. yoelii (17XL), transgenic mice had a lower peak parasitemia, cleared their infection by day 24, and survived; control mice died between days 11 and 13. In transgenic mice with a P. chabaudi adami infection, the parasite parasitemia rose quickly but was also rapidly cleared. The infection was not reversed by splenectomy. In biochemical studies, Shear et al. also showed that human HbF was digested half as well as human HbA by a recombinant parasite hemoglobinase (Plasmepsin II) and proposed stability of the HbF tetramer as a potential mechanism for protection by HbF.

The differential growth in the two media found by Friedman and the lack of specific details regarding the RBCs of the transgenic mice and their relevance to polymorphic human populations, as well as the purely biochemical data collected in Shear et al., still left questions regarding HbF's detrimental influence on growth. Therefore in 2011, Amaratunga et al. re-examined this finding and showed no difference in development in cultures of six cord blood samples and one HPFH compared to HbA dominant blood using five different parasite lines (7G8, FVO, GB4, MC/R+, TM284). Flow cytometry was employed to assess development, measuring DNA replication in growth, compared to the previous work that relied on subjective morphological assessment. Amaratunga and colleagues however did not assess for age of RBCs, nor did they follow cultures beyond 40 h (a noted limitation due to known RBC aging observed in in vitro culture) so a growth delay from the trophozoite to the schizont stage was still possible. Amaratunga et al. also demonstrated a reduction in PfEMP-1 surface expression, rosetting, and adherence to microvascular endothelial cells and monocytes by parasitized cord and HPFH RBCs compared to HbA dominant RBCs. They proposed a new protective mechanism of HbF, namely reduced cytoadherence.¹²

Considering the finding of no difference in parasite growth in cord and HbA dominant RBCs by Amaratunga et al., Sauerzopf et al. investigated the potential role of maternal immunoglobulins in limiting growth.¹³ They too found no difference in parasite growth in cord and HbA dominant RBCs using the HRP2-assay. However, when non-malaria-exposed donor plasma was used compared to maternal or cord blood plasma (presumably exposed); growth in the plasma from malaria exposed individuals was reduced. Cyrklaff et al. also found no effect on parasite multiplication rate and hemozoin levels up to 12 cycles but demonstrated impaired actin remodeling in HbF RBCs¹⁴ by comparing cryo-tomographic images of FCR3^CSA in adult, cord, sickle cell trait, and hemoglobin C trait erythrocytes. The shorter actin filaments in non-HbAA RBCs suggested a dysfunctional cytoskeleton. In addition, Maurer's clefts and knobs were malformed and the amount of VAR2CSA presented on the red cell surface was reduced. These later studies all pointed toward reduced cytoadherence and not growth inhibition as a mechanism of HbF protection.

To elucidate the potential effect of HbF in patients with sickle cell disease treated with hydroxyurea, Archer et al. also compared overall parasite growth and proliferation in umbilical cord blood and heterozygote African HPFH versus adult blood (HbAA only) using cell sorting and the cultured parasite line Pf3D7 IG06. No significant difference in parasite intraerythrocytic development and parasite multiplication rate (PMR) was noted in HbF and HbA dominant RBCs,¹⁵ in contrast with the significant growth defect seen in sickle cell trait RBCs in low oxygen.⁶

In reviewing these various data, the authors, recommend the following when considering future studies of HbF's effect on malaria invasion, development, and proliferation: (1) RBC age is likely a confounder given the numerous changes experienced by the RBC during erythropoiesis. To avoid age differences in RBC, HPFH in the absence of other hemoglobinopathies or isogenic RBC lines should be used; (2) While cultured parasites are easier to obtain and manipulate, where possible, primary field isolates should be used to confirm findings; (3) Given the polygenicity of HbF with other genes,^{16, 17} it is important that large sample sizes be used to establish small size effects. When the number of RBC blood samples is finite, single cell isolation of HbF positive and negative cells as done by Pasvol et al. via acid-elution and more recently by flow cytometry should be employed.

In summary, the effect of HbF on parasite growth is at best relatively small and may only be significant in the setting of a high oxidant stress environment. However, in the context of polygenic effects of protection against lethal malarial infection, HbF might play a role for better or for worse. Future studies should consider these possibilities.

人类疟原虫恶性疟原虫与其宿主红细胞 (RBC) 中的众多多态性分子相互作用，这给检查单一 RBC 决定簇对寄生虫感染的影响带来了困难。胎儿血红蛋白 (HbF) 的故事就是一个很好的例子。两位作者在大约 50 年前首次探讨了 HbF 是否抑制侵袭和/或生长的问题，因为流行病学数据的获取表明 HbF 和镰状血红蛋白 (HbS) 之间的相互作用可能会影响疟疾感染。^{1, 2}在这里，我们详细介绍了调查 HbF 对疟疾感染影响的研究，强调了与解释相关的复杂性，并就如何进一步研究自然群体中寄生红细胞的多态性提供了建议。

在使用受感染的混合脐带和 HbAA 细胞体外实验培养的基础研究中，Pasvol 等人。表明恶性疟原虫偏爱年轻的红细胞。³这是通过分析患者的血涂片和经过改良的 Kleihauer 等人的酸洗脱差异染色方法处理的培养物来完成的，该方法可将寄生的 HbF 细胞 (F 细胞) 与正常 (HbAA) 细胞区分开来，因为HbF 在 pH 3.5 下可抵抗洗脱。⁴因此，可以在相同条件下而不是在平行培养物中测试具有不同血红蛋白的细胞。脐带和 HbAA 血液混合物中 F 细胞（即更年轻、代谢更活跃的细胞）比例较高，被寄生虫入侵。此外，在所使用的单一脐带血样本的 F 细胞中发现了较年轻阶段的寄生虫。与镰状细胞性状红细胞生长完全停滞相比，^{5、6 的}生长延迟轻微但显着。最后，帕斯沃尔等人。³证实恶性疟原虫优先侵入婴儿血液样本中较年轻的红细胞，而不是 F 细胞。婴儿血液中较年轻的细胞，即那些以 HbA 为主的细胞，在最多两个细胞周期中更容易受到侵袭。虽然这些研究清楚地证明了年龄而非 HbF 对寄生虫入侵的影响，但注意到 HbF 对生长有轻微但可测量的影响。

随后的一项使用 HbF 和 HbA 占主导地位的红细胞混合物的研究再次证实，入侵不受 HbF 的影响，而是受红细胞年龄的影响。⁷除上述实验中使用的脐带-成人混合血和 3-6 个月婴儿的血液外，还侵入了非洲遗传性胎儿血红蛋白持久性 (HPFH) 杂合子血液和英国 HPFH 纯合子血液的混合物和成人 HbAA 血液进行了研究。前两个样本中较年轻的细胞（脐带和 HbAA 红细胞）再次优先受到侵袭。在两个 HPFH 样本中，所有红细胞的年龄相同，不存在对 HbF 或 HbA 占主导地位的红细胞的偏好。帕斯沃尔、韦瑟罗尔和威尔逊还发表了另一组研究，强调恶性疟原虫优先入侵网织红细胞。⁸在这些实验中，离心后从 Wintrobe 管的顶部、中部和底部分离红细胞。然后用亮甲酚蓝对红细胞进行网织红细胞染色，并用吉姆萨染色剂对寄生虫进行复染。与正常细胞相比，在网织红细胞和含有网织红细胞的顶部部分中发现了更高的寄生虫血症。威尔逊等人。还表明，来自脐带血和成人血的单一培养物的红细胞中的寄生虫大小在 F 细胞中更小。⁹这些研究再次表明，HbF 对侵袭没有影响，但对生长有影响。

在这些早期研究中，重点主要是年轻细胞的优先寄生虫入侵，并对 F 细胞中寄生虫发育延迟进行了简短评论。因此，其他研究^10、11特别关注 HbF 诱导发育迟缓的机制。弗里德曼等人。只能在体外用谷胱甘肽还原培养基（培养基 199）（与 Pasvol 和 Wilson 研究中使用的培养基相同）显示对 HbF 占主导地位的红细胞（未指定的红细胞类型）中寄生虫生长的抑制。在RPMI（罗斯威尔公园纪念研究所）培养基中，观察到正常生长。这表明了氧化应激的机制。¹⁰后来，Shear 等人在具有人源化 α 和 γ 珠蛋白链的转基因啮齿动物疟疾模型中证明了 HbF 的体内保护作用。¹¹当感染约氏疟原虫(17XL) 时，转基因小鼠的寄生虫血症峰值较低，并在第 24 天清除感染并存活下来；对照小鼠在第 11 天和第 13 天之间死亡。在感染P. chabaudi adami 的转基因小鼠中，寄生虫血症迅速上升，但也很快被清除。脾切除术未能逆转感染。在生化研究中，Shear 等人。还表明，重组寄生虫血红蛋白酶（Plasmepsin II）对人 HbF 的消化率是人 HbA 的一半，并提出 HbF 四聚体的稳定性作为 HbF 保护的潜在机制。

Friedman 发现的两种培养基中的生长差异，以及缺乏有关转基因小鼠红细胞及其与多态人类群体的相关性的具体细节，以及 Shear 等人收集的纯生化数据，仍然对 HbF 的问题留下了疑问。对生长产生不利影响。因此，2011 年，Amaratunga 等人。重新检验了这一发现，结果显示，使用五种不同的寄生虫系（7G8、FVO、GB4、MC/R+、TM284），与 HbA 优势血相比，六份脐带血样本和一份 HPFH 的培养物的发育没有差异。与之前依赖主观形态学评估的工作相比，流式细胞术被用来评估发育，测量生长过程中的 DNA 复制。然而，Amaratunga 和同事没有评估红细胞的年龄，也没有跟踪超过 40 小时的培养物（这是由于体外培养中观察到的已知红细胞老化而引起的明显限制），因此从滋养体到裂殖体阶段的生长延迟仍然是可能的。阿马拉通加等人。还证明，与 HbA 占主导地位的红细胞相比，寄生的脐带和 HPFH 红细胞的 PfEMP-1 表面表达、玫瑰花结以及对微血管内皮细胞和单核细胞的粘附减少。他们提出了 HbF 的一种新的保护机制，即减少细胞粘附。¹²

考虑到 Amaratunga 等人发现脐带和 HbA 占主导地位的红细胞中寄生虫生长没有差异，Sauerzopf 等人。研究了母体免疫球蛋白在限制生长方面的潜在作用。¹³他们使用 HRP2 检测也发现脐带血和 HbA 占主导地位的红细胞中的寄生虫生长没有差异。然而，当使用未暴露于疟疾的供体血浆与母体或脐带血浆（可能暴露）进行比较时；疟疾暴露者的血浆生长减少。赛克拉夫等人。还发现长达 12 个周期对寄生虫增殖率和疟原虫色素水平没有影响，但通过比较成人、脐带、镰状细胞特征和血红蛋白 C 特征红细胞中FCR3 ^CSA的冷冻断层扫描图像，证明 HbF RBC ^{14中的肌动蛋白重塑受损。}非 HbAA 红细胞中较短的肌动蛋白丝表明细胞骨架功能失调。此外，Maurer 裂和结畸形，红细胞表面存在的 VAR2CSA 量减少。这些后来的研究都指出减少细胞粘附而不是生长抑制作为 HbF 保护的机制。

为了阐明 HbF 对接受羟基脲治疗的镰状细胞病患者的潜在影响，Archer 等人。还使用细胞分选和培养的寄生虫系 Pf3D7 IG06 比较了脐带血和杂合子非洲 HPFH 与成人血液（仅 HbAA）中寄生虫的总体生长和增殖。在 HbF 和 HbA 占主导地位的红细胞中，寄生虫红细胞内发育和寄生虫增殖率 (PMR) 没有显着差异，¹⁵与低氧条件下镰状细胞性状红细胞中观察到的显着生长缺陷形成鲜明对比。⁶

在回顾这些不同的数据时，作者在考虑未来研究 HbF 对疟疾侵袭、发展和增殖的影响时提出以下建议：(1) 鉴于红细胞在红细胞生成过程中经历的众多变化，红细胞年龄可能是一个混杂因素。为了避免红细胞年龄差异，应使用没有其他血红蛋白病或同基因红细胞系的 HPFH；(2) 虽然培养的寄生虫更容易获得和操作，但在可能的情况下，应使用初级现场分离株来证实发现；(3) 鉴于 HbF 与其他基因的多基因性，^{16, 17}使用大样本量来建立小样本效应非常重要。当 RBC 血液样本数量有限时，如 Pasvol 等人所做的那样，对 HbF 阳性和阴性细胞进行单细胞分离。应采用通过酸洗脱和最近的流式细胞术。

总之，HbF 对寄生虫生长的影响充其量是相对较小的，并且可能仅在高氧化应激环境中才显着。然而，在防止致命性疟疾感染的多基因效应的背景下，HbF 可能发挥作用，无论好坏。未来的研究应该考虑这些可能性。